Input
$0.10 / MTok for prompts up to 100,000 tokens
$0.50 / MTok for prompts over 100,000 tokens
Output
$0.50 / MTok for prompts up to 100,000 tokens
$2.50 / MTok for prompts over 100,000 tokens
100k tokens is an absurdly low cutoff and it is only applicable to Haiku and not Sonnet or Opus. It's a low enough cutoff that it will be quickly exceeded if you are doing anything with Agents; for typical generation or Jev-like classifiers, it's a good value and as noted in this article, that is apparently the vast majority of Haiku use.
In both cases, still much cheaper than Haiku 4.5's $1 input / $5 output and these prices better compete with GPT-6 Luna. ($0.10 input / $0.50 output, but with no token threshold [EDIT: the threshold for Luna is apparently 272k])
Haiku 5.5 is noticeably smarter than GPT-6 Luna, so I can see their pricing strategy here.
For a while Anthropic has lacked a cost effective “cheap” LLM for summarisation, compacting, RAG helpers, etc.
These ‘ephemeral’ workloads are often under 100k tokens, or can be structured to be under 100k.
In some coding benchmarks, Haiku 5.5 beats Sonnet 5! (Especially implementation; do a well defined Jira ticket; etc), it’s really impressive how much intelligence per dollar has grown in just a few short months.
There's also a tokenizer efficiency difference: modern Claude's 100K tokens are about ~60-65K modern GPT tokens, so in reality the Luna cutoff is much further away than the Haiku one.
You could also use it as a subagent prompted eg by Sonnet/Opus orchestrator agent and for many agentic workflows significant part of the dispatched tasks might be under 100k budget.
Notable that one suggested use case for Haiku is "classification requests", i.e. Jev competitor, and the pricing matches GPT-6 Luna which is behind OpenAI's "Decisions API" Jev competitor.
For this application 100K token input is plenty.
Of course Anthropic and OpenAI, both at $0.10/M, are still 2.5x the cost of Jev's $0.04/M.
The classification performance remains to be seen, but presumably we'll soon start to see classification benchmarks.
For other tasks like summaries (another suggested usage) it's good to see Luna and Haiku now competing against each other on cost.
I'd love to know how the business automation market breaks down by volume of call type though - hard to imagine that decision making (e.g. branching, triage) isn't a very large part of it, greater than these other suggested Haiku use cases.
So even at the 1.5x/2x rate luna is still half the price of this. Weird pricing strategy from Anthropic. I'm sticking with Luna if I don't need a super smart model
They are targeting businesses/API use for fast decision making and agent integration. Plus they now need to be competitive with Jev-type models in that space.
I with they'd give Haiku like 400k tokens roughly, I think between 400k or even 600k tokens is a sweet spot, but Haiku is basically designed to be for small edits is my understanding, but it sucks because any time I ask Opus to "try" letting Haiku do the work, it just falls apart and Opus comes back and tells me it switched to Sonnet (even before Sonnet finally jumped up to 5.x).
I will try the new Haiku, but it would be worthwhile if Haiku could take sane instructions and do all file editing for Opus / Sonnet / Fable then it would be worth using.
I mostly use Haiku for really, really basic stuff, never for actual engaging work. I've used it for first-pass analysis to triage bugs, for example - all it does is related N bugs together to see if any potentially relate. Then I have Sonnet investigate further.
>> 100k tokens is an absurdly low cutoff and it is only applicable to Haiku and not Sonnet or Opus. It's a low enough cutoff that it will be quickly exceeded if you are doing anything with Agents
Your vibes don't appear to be supported by facts. From the announcement:
>> Claude Haiku 5.5 is priced 90% lower than Claude Haiku 4.5 for requests up to 100,000 tokens, and 50% lower for requests over 100,000 tokens. On Haiku 4.5, 90% of requests fell into the former category.
Well, unless you're using OpenCode Go, it's per-token costs (even if already super low), while Haiku falls under the Claude sub. It's just more straight forward and you aren't feeling a "loss" with the sub.
That's not what any benchmarks that look at cost per task or similar says in terms of cost. The Chinese models, generally speaking, might be cheaper per token but need a lot more tokens to get there.
Some people/organizations are ideologically opposed to using Chinese models. Not me, I use GLM-5.3-Flash for almost everything (the subscription-subsidized pricing on a legacy Z.ai plan makes it the best value model by a wide margin), along with some MiMo and DeepSeek. Still, I use Luna for certain tasks where speed is more valuable than performance; I can see this new Haiku displacing Luna for those. If you mean Haiku 4.5 though I agree, that model was a waste of time and money.
> Second, this week, we’ll roll out a new monthly API credit to all Max and Team subscribers for use on the Claude Platform. Max 5x users will get $100 in credits per month, Max 20x users will get $200, and Team subscribers will receive up to $500, pooled across their users
This is a very big benefit for me. I can now ship actual ai enhanced features behind my subscription without paying extra or fully relying on on-device models. I do worry that this is to soften the blow for user-unfriendly changes
This is them sneaking in taking the Claude Agent SDK off of subscription plans through the back door along with a model release. They previously wanted to do this in June, but backpedaled after huge backlash:
Those mfers. I'm using this for work! I use my work teams plan with pi so I can do all kinds of custom workflows that I can't in Claude Code. Time to convince management I need OpenAI instead.
Even after the June changes there was some allowance to use agent SDK on the pro plan. This will move me to codex tomorrow if agent SDK is really blocked on pro
This is massive. So on top of the regular usage, we now have USD 200,- to freely use via the API however we please, even resell? That is a statement, even knowing that inference does not cost them nearly as much as they charge, this is very developer-friendly. Does some minor de-risking for testing concepts. Terms seem to be reasonable [0].
Of course, they don't do this out of pure kindness, but I really struggle to see a negative for subscribers, especially given changing to another model is essentially frictionless, so once the monthly allowance is used up, you can still just decide not to use Anthropic models for the remainder.
Nah, it's pretty trivial to switch providers (especially with Claude's help, ha).
This is more to encourage people to try out adding AI into their product, which is a totally different flow and experience from using AI to build the product.
Not really, you have to fiddle with generating api keys and setting environment variables. Meanwhile with Anthropic it will just start charging you API prices for the tokens you are generating without even a single warning.
One interesting thing is it took a look at the job at hand, and immediately delegated it to Opus 5.5. It at least knows what it isn't good at. Very fast though, and likely best used for small subagent tasks / tightly scoped work.
Neither of these look "good" to me. There is so much visual noise on the page, like someone turned the "AI Slop" dial to 11. In fact I prefer the simpler design Haiku made.
It's not really about whether the design looks good. It's about if the model can take the design given to it and replicate it in code. Opus 5.5 matches the designs almost to the pixel. Haiku built something else entirely.
Totally fair, but I'd encourage you not to look at the design so much as the task. This was a design that's part of a benchmark test suite specifically for image->html conversion. The dense visual noise / complexity / flowing svg shapes are things that most LLMs have trouble with.
At work we use haiku 4.5 for a handful of latency sensitive tasks that are fairly simple. It performs well. Just started testing 5.5 as I’ve been anticipating a nice improvement since it was teased. Results so far are trash. Prompt leakage even. And it’s slower. I guess it’s cheap but I think they got the balance wrong on this.
The monthly API credits for Max plan seems fantastic, especially considering Haiku pricing. Being able to actually use my Claude plan for other harnesses and use-cases on top of regular CC usage is everything I wanted.
Anthropic has really been doing all the right things in the past few weeks, while OpenAI continues to fumble the bag.
Note that this is Anthropic Trojan-Horsing the previously announced June change in with a model release, where the Claude Agent SDK can no longer be used with Claude subscriptions and is now billed with API credits only.
yeah totally agree. esp how efficient it can be to have a subscription quota-paid orch spin up a bunch of API agents, this is kind of like free money to encourage what was already an easy way to save money (via batch pricing)
This is great! Been using GPT 6 Luna for decompiling my childhood favorite game (Age of Mythology) and this means I can throw Haiku into the mix as well. 17352/21965 functions matched so far...
And to setup a harness that will decompile the game and start doing a matching decompilation of every function. It set up a bunch of tooling and started a service in the background to do this actual decompilation campaign. I put some instructions into the main opus chat now and then to e.g. add automatic git pushing including a nice svg chart of progress and to switch model strategies here and there i.e. to do a first pass with a cheap model and then switch to opus/sol if the small model can't solve it.
About time Anthropic released a competitive cheap model. Haiku 4.5 has been too expensive compared to its performance for months now (in fact I don't remember being too impressed even when it was released). This one actually looks worth using in some scenarios. If it's really as much of a step up from Luna as the benchmarks they've shown indicate, it'll probably replace Luna in my workflows. 100k tokens is a pretty low threshold before the price goes up, but I tend to use these smaller models for smaller tasks anyway.
> Claude Haiku 5.5 is our fastest model to date at each model’s standard speed, although it runs less quickly than our Opus models in Fast Mode.
Opus 5.5 runs 117 tps average on Openrouter, so it must be at least 10-20 tps slower for them to mention. IDK why they mention this as it does not help for marketing though.
https://openrouter.ai/anthropic/claude-opus-5.5
131tok/s P50 according to OpenRouter currently, though might move up or down over the coming days. If it sticks at that speed, roughly twice the throughput of Luna is impressive, though the 5x price increase beyond 100k is painful.
Was a big fan of Haiku 4.5, though understand why for most Sonnet was the far better option.
Yes -> every 18 months they've gotten 90% more efficient for the same level of quality for about 5 years. There's little sign that trend is slowing. If anything, there's reason to believe that System 1 models (plus potentially 1-2-3 workflows) may increase that over the next 3-5 years.
You'll know when the trend stops -> when the intelligence differential between smaller models like 7B starts to grow instead of shrink from 32B models -> that means 7B is getting about as smart as it can get. Then, 32B will follow next, then 70B, etc etc.
Super intelligence that doesn't have to deal with the real world, maybe.
I wouldn't be surprised if less than 1B param equivalent of our brain deals with solving math and writing computer programs and physics and all the things we tend to associate with "intelligence" - especially if you ultra optimized for that, I doubt our brain works like that.
Dealing with the real world, I highly highly doubt it.
How about if we get away from written text as the input, to something more fundamental, that then also is able to produce text (among other things)?
Given that humans learn to talk while having encountered a measly number of word instances, and, given enough time, we should always be able to improve on the lottery that is biology, it does seems fairly likely.
I've heard tell about 100% of certain types of work being ended in batches of six months. For years. Truthfully, I'm skeptical, but accuracy wasn't prioritized.
serious bit: if you think about how these smaller models work, at the end of the day it seems that they are now capable of forgetting useless information because they're able to derive it in reasoning allowing models to become smaller at the cost of requiring more reasoning tokens to solve a task.
Knowledge will be shifted to systems like n-gram augmentation which are relatively cheap and will not compete with reasoning capabilities for weight saturation.
Apparently quite a bit smarter than Luna, I wonder what use cases it can cover. I actually honestly don't need a Haiku level AI to be that smart, and looks like you pay for it in the per token cost, I need speed mainly. I might even rather have a dumber but much faster model for things like web searching and parsing to retrieve results for the app or other LLM to do things with.
Yeah, people like to poop on the pelican. But pelican quality still correlated with overall model capabilities reasonably well, and you can immediately see and interpret it. It's a running gag, but it also does have some actual value.
Both you and the OP are asking the wrong questions. You are requesting a benchmark that tests for nothing.
Lets not focus on slop benchmarks and instead focus on more rigorous ones such as KernelBench [0], CompileBench [1], TraderBench [2] and Harvey's Legal Agent Benchmark [3] as frontier-level benchmarks that are far better to verify whilst requiring expert level intelligence as the barrier.
I've been using GPT-6 Luna in some capacity for nearly all my agent workflows. It's just a really good model, and the pricing is cheap. If Haiku 5.5 is better, and the same price (under 100k context... which is a big caveat) i'd probably swap it.
It’s absolutely better than Luna. It feels closer to a “sonnet 5.2” if that makes sense.
Of course it’s not as big, and hence falls-off quicker. I’d consider the 100k a “promotional price” to match Luna’s token pricing while delivering noticeably more intelligence.
I have a zsh functions that calls claude code with haiku to suggest commit messages, is faster and the instructions are two lines.
I also have an "ask" script that I use daily to ask simple stuff, it can access websearch and webfetch, it's more than enough to parse logs, ask for commands, quick research on the internet, small stuff. https://github.com/mariocesar/dotfiles/blob/main/common/.loc...
I use haiku for things that needs to be quick, have really clear instructions.
Opus often picks it when it's doing a "find me something" subagent. But largely it's been held back by being fully a year old at this point, and priced at a much higher price than models that are far more capable.
not haiku, but luna - last week i used it for things like "read this historical dump of 15k support tickets and break them into categories that make sense, then propose help docs that i could write to handle the most frequent queries in each category"
It's great at parsing documents inexpensively. For the few skills/plugins I've made, I usually instruct Claude to use Haiku for low-reasoning grunt work.
The forgotten model is back on the map. I actually got OK mileage when I tried it for coding months ago. Maybe I'll try it again, with Opus guiding it, and see how it goes.
IMO, this is better. Luna is super cheap, but it's not that capable. At higher levels of reasoning, it's not that fast.
This is more expensive, but it also looks like it's better enough that it's far more useful.
I also won't be surprised if you look at cost per completed task + wall clock time that it comes out ahead for the majority of what you'd want to actually use it for.
Luna will still be a great option for doing non-engineering tasks super cheaply.
That was effectively required to match GPT-6.1 Sol (costs and caching prices are now equal). Sonnet 5.5 made zero sense to use over Opus 5.5 under the old cache prices.
> Second, this week, we’ll roll out a new monthly API credit to all Max and Team subscribers for use on the Claude Platform. Max 5x users will get $100 in credits per month, Max 20x users will get $200, and Team subscribers will receive up to $500, pooled across their users. These credits are designed to allow our users to experiment with building tools, apps, and agents that call our API. They can be used on any of our models. For more information, see our Help Center article.
Did anyone read this? We get free API credits on some plans now
> Haiku 5.5’s cybersecurity safeguards are more restrictive than Haiku 4.5’s, but somewhat less restrictive than those we’ve applied to other recent models. In cybersecurity, they permit a wider range of defensive tasks than our safeguards for Sonnet 5.5, but they still block penetration testing and other techniques more likely to be used by attackers.
If you block pentest or "other techniques more likely to be used by attackers", then what does "permit a wider range of defensive tasks" even mean?
Any defensive task that's meaningful is almost indistinguishable from legitimate red-teaming that then falls under 'likely to be used by attackers". If only they would just stop nerfing these models, that'd be great. No APT is waiting around for Anthropic's permission, so might as well let us have some cool stuff.
> but they still block penetration testing and other techniques more likely to be used by attackers.
>
> Haiku 5.5’s biology safeguards are the same as for Sonnet 5, Sonnet 5.5, and Opus 5. They allow research biology questions but restrict access to requests that we judge as likely to cause harm. Organizations working on wider-ranging biology and cyber activities can apply to our Life Sciences Verification Program and Cyber Verification Program.
I would like to take a moment of your time to tell you about some of the "bioweapons" Anthropic has blocked that involved Haiku!
> Importantly, because our biological safety classifiers robustly block content involving high-risk biological research (in this case, the construction of enhanced pandemic potential pathogens), all of these exchanges occurred on models in our weakest class of models (specifically, the models were Claude Sonnet 4 and Haiku 4.5, the latter of which the user began using after Sonnet 4 was deprecated).
>
> Upon a detailed examination of the exchanges, we estimate that the uplift provided by Claude was primarily clerical assistance in data analysis, study ideation and design. This is consistent with our understanding of the capabilities of Sonnet 4 and Haiku 4.5, which are not able to perform expert-level biology research tasks; we estimate that the uplift provided to the researcher was limited and substantially lower than it would have been from one of our more capable models.
Anthropic then says for the above, "we estimate that the uplift provided by Claude was primarily clerical assistance in data analysis, study ideation and design"
While doing my best to avoid comment, please note, they're talking about a domain expert in a state research institution using Claude to do paperwork.
What did they save us from? What bioweapons did these filters prevent? From the front matter report,
> The above LLM platform is not the only route via which researchers engaged in viral gain-of-function research have used our platform. In May 2026, we discovered a researcher outside the US using Claude in their research on highly-pathogenic avian influenza (“bird flu”). The research focused on viruses’ adaptation to mammals, and the mechanism by which it causes severe disease beyond the respiratory tract.
OK. Sounds serious. "Gain of function research..." but who and why?
> The researcher pursued this work in a credible institutional context, and interacted with Claude over the course of several weeks, exchanging thousands of messages. In these exchanges, the researcher leveraged Claude’s knowledge of the scientific literature to assist the researcher in study planning and design, data analysis, and the interpretation and prioritization of experiments. The researcher also used Claude for editorial assistance in writing up the research.
So this was a researcher inside of some country's national lab ("credible institutional context") doing research on dangerous viruses using Claude for "for editorial assistance in writing up the research."
What "uplift" are you providing to scientists working at specialized global BSL-4 labs that already have – and I quote their report - "physical access to such isolates." (as in samples of viruses)?
This nonsense has been expanded with even worse "safeguards."
This makes Claude unusable for any serious scientist or people curious about science, which is sad.
Pricing is...a bit weird.
100k tokens is an absurdly low cutoff and it is only applicable to Haiku and not Sonnet or Opus. It's a low enough cutoff that it will be quickly exceeded if you are doing anything with Agents; for typical generation or Jev-like classifiers, it's a good value and as noted in this article, that is apparently the vast majority of Haiku use.In both cases, still much cheaper than Haiku 4.5's $1 input / $5 output and these prices better compete with GPT-6 Luna. ($0.10 input / $0.50 output, but with no token threshold [EDIT: the threshold for Luna is apparently 272k])
Haiku 5.5 is noticeably smarter than GPT-6 Luna, so I can see their pricing strategy here.
For a while Anthropic has lacked a cost effective “cheap” LLM for summarisation, compacting, RAG helpers, etc.
These ‘ephemeral’ workloads are often under 100k tokens, or can be structured to be under 100k.
In some coding benchmarks, Haiku 5.5 beats Sonnet 5! (Especially implementation; do a well defined Jira ticket; etc), it’s really impressive how much intelligence per dollar has grown in just a few short months.
There's also a tokenizer efficiency difference: modern Claude's 100K tokens are about ~60-65K modern GPT tokens, so in reality the Luna cutoff is much further away than the Haiku one.
You can test with Anthropic's count_tokens endpoint or with https://crates.io/crates/tokwc
> ...this tokenizer, the same input text produces approximately 30% more tokens on Claude Haiku 5.5 than on Claude Haiku 4.5.
So, it is might be even worse.
No, it's just Haiku 4.5 is so old that it predates the new Claude tokenizer change in Claude 4.7+
It's actually existing flat per-token pricing that is weird.
Neither encode nor decode are linear in compute, so providers need to price for average expected length.
This is just getting closer to the true cost of generating tokens.
You could also use it as a subagent prompted eg by Sonnet/Opus orchestrator agent and for many agentic workflows significant part of the dispatched tasks might be under 100k budget.
Notable that one suggested use case for Haiku is "classification requests", i.e. Jev competitor, and the pricing matches GPT-6 Luna which is behind OpenAI's "Decisions API" Jev competitor.
For this application 100K token input is plenty.
Of course Anthropic and OpenAI, both at $0.10/M, are still 2.5x the cost of Jev's $0.04/M.
I think the 2.5 times cost but actually pays off in terms of intelligence compared to jev and the general capability of using it beyond classification
The classification performance remains to be seen, but presumably we'll soon start to see classification benchmarks.
For other tasks like summaries (another suggested usage) it's good to see Luna and Haiku now competing against each other on cost.
I'd love to know how the business automation market breaks down by volume of call type though - hard to imagine that decision making (e.g. branching, triage) isn't a very large part of it, greater than these other suggested Haiku use cases.
Luna does as well, but just at a higher limit.
From OpenAI's website: Prompts with more than 272K input tokens are priced at 2x input and cache rates and 1.5x output for the full request.
So even at the 1.5x/2x rate luna is still half the price of this. Weird pricing strategy from Anthropic. I'm sticking with Luna if I don't need a super smart model
Huh, that disclaimer is on the model page (https://developers.openai.com/api/docs/models/gpt-6-luna) but not the pricing page. Annoying.
Fixed.
They are targeting businesses/API use for fast decision making and agent integration. Plus they now need to be competitive with Jev-type models in that space.
I with they'd give Haiku like 400k tokens roughly, I think between 400k or even 600k tokens is a sweet spot, but Haiku is basically designed to be for small edits is my understanding, but it sucks because any time I ask Opus to "try" letting Haiku do the work, it just falls apart and Opus comes back and tells me it switched to Sonnet (even before Sonnet finally jumped up to 5.x).
I will try the new Haiku, but it would be worthwhile if Haiku could take sane instructions and do all file editing for Opus / Sonnet / Fable then it would be worth using.
I mostly use Haiku for really, really basic stuff, never for actual engaging work. I've used it for first-pass analysis to triage bugs, for example - all it does is related N bugs together to see if any potentially relate. Then I have Sonnet investigate further.
>> 100k tokens is an absurdly low cutoff and it is only applicable to Haiku and not Sonnet or Opus. It's a low enough cutoff that it will be quickly exceeded if you are doing anything with Agents
Your vibes don't appear to be supported by facts. From the announcement:
>> Claude Haiku 5.5 is priced 90% lower than Claude Haiku 4.5 for requests up to 100,000 tokens, and 50% lower for requests over 100,000 tokens. On Haiku 4.5, 90% of requests fell into the former category.
Haiku 4.5 users were using it for Kleenex requests because that was the best it could do.
People weren't using Haiku 4.5 for agents before. 5.5 is good enough that it might be.
encode and decode tok/s which is ($/s) when it comes to pricing drops heavily above 100k tokens.
There are plenty of workflows like translations where you'd easily be under the cap.
It's their creative way of 'matching' Luna's prices.
It could be to incentivize people to not be lazy users of tokens.
Who in their right mind would use haiku while Mimo or GLM cost 10% of what they are charging with much smarter models?
Well, unless you're using OpenCode Go, it's per-token costs (even if already super low), while Haiku falls under the Claude sub. It's just more straight forward and you aren't feeling a "loss" with the sub.
That's not what any benchmarks that look at cost per task or similar says in terms of cost. The Chinese models, generally speaking, might be cheaper per token but need a lot more tokens to get there.
Some people/organizations are ideologically opposed to using Chinese models. Not me, I use GLM-5.3-Flash for almost everything (the subscription-subsidized pricing on a legacy Z.ai plan makes it the best value model by a wide margin), along with some MiMo and DeepSeek. Still, I use Luna for certain tasks where speed is more valuable than performance; I can see this new Haiku displacing Luna for those. If you mean Haiku 4.5 though I agree, that model was a waste of time and money.
There really aren't any models at 10% of the price of Luna or Haiku.
Presumably everyone who doesn't bother integrating a third party API key into their harness, which would probably be most of the Claude Code users.
> Second, this week, we’ll roll out a new monthly API credit to all Max and Team subscribers for use on the Claude Platform. Max 5x users will get $100 in credits per month, Max 20x users will get $200, and Team subscribers will receive up to $500, pooled across their users
This is a very big benefit for me. I can now ship actual ai enhanced features behind my subscription without paying extra or fully relying on on-device models. I do worry that this is to soften the blow for user-unfriendly changes
This is them sneaking in taking the Claude Agent SDK off of subscription plans through the back door along with a model release. They previously wanted to do this in June, but backpedaled after huge backlash:
https://support.claude.com/en/articles/15036540-use-the-clau...
Those mfers. I'm using this for work! I use my work teams plan with pi so I can do all kinds of custom workflows that I can't in Claude Code. Time to convince management I need OpenAI instead.
Do we know if claude -p is now drawing from this API usage?
Even after the June changes there was some allowance to use agent SDK on the pro plan. This will move me to codex tomorrow if agent SDK is really blocked on pro
This is massive. So on top of the regular usage, we now have USD 200,- to freely use via the API however we please, even resell? That is a statement, even knowing that inference does not cost them nearly as much as they charge, this is very developer-friendly. Does some minor de-risking for testing concepts. Terms seem to be reasonable [0].
Of course, they don't do this out of pure kindness, but I really struggle to see a negative for subscribers, especially given changing to another model is essentially frictionless, so once the monthly allowance is used up, you can still just decide not to use Anthropic models for the remainder.
[0] https://www.anthropic.com/legal/credit-terms
This is literally for you to get tangled in their api and when they stop giving you the allowance they hope you will just continue to pay
Nah, it's pretty trivial to switch providers (especially with Claude's help, ha).
This is more to encourage people to try out adding AI into their product, which is a totally different flow and experience from using AI to build the product.
That's an old tactic for an old world. You only need, what, half an hour with your agent of choice to write you out of that?
What does "tangled in their api" mean? Switching is pretty easy.
Not really, you have to fiddle with generating api keys and setting environment variables. Meanwhile with Anthropic it will just start charging you API prices for the tokens you are generating without even a single warning.
>> Not really, you have to fiddle with generating api keys and setting environment variables.
That's 5-15 minutes of work at most. Not exactly the type of lock-in the parent is implying.
The user could have always done that regardless of if the user has the option to be charged API rates on or off.
OpenAI will probably add this to their plans within a week
With OpenAI you can just use Oauth and get a token to use your subscription.
Anthropic isn't even close to being this useful.
Biggest reason for an OAI subscription instead of Ant imo.
Biggest loss is that Ant models look like they are genuinely better.
> Biggest loss is that Ant models look like they are genuinely better.
This changes on a weekly basis, I ended up with subscriptions to most of the providers (except for X.ai).
Ran image -> html tests for this. I was curious if this smaller model was good enough for complex UI. It was not.
Haiku 5.5: https://html.non.io/lcars-haiku-5.5/
Opus 5.5 for comparison: https://html.non.io/lcars-opus-5.5
Designs it was building from: https://diffui.ai/app/canvas/5093e689-1e74-4f26-b632-2a4500f...
One interesting thing is it took a look at the job at hand, and immediately delegated it to Opus 5.5. It at least knows what it isn't good at. Very fast though, and likely best used for small subagent tasks / tightly scoped work.
Pac-Man Bench:
Considering the price, no model comes close to being as good as this. However, it did take an extremely long time.
TIME 19m COST $0.16 https://jonclegg.github.io/pacman-bakeoff/#claude-haiku-5-5
All results: https://jonclegg.github.io/pacman-bakeoff/
To be fair, you're making it compete with the best public LLM right now that's 2 size/price tiers above it.
Neither of these look "good" to me. There is so much visual noise on the page, like someone turned the "AI Slop" dial to 11. In fact I prefer the simpler design Haiku made.
It's not really about whether the design looks good. It's about if the model can take the design given to it and replicate it in code. Opus 5.5 matches the designs almost to the pixel. Haiku built something else entirely.
I guess I'm giving GP feedback about their product diffui.ai, not really about Opus' performance.
Totally fair, but I'd encourage you not to look at the design so much as the task. This was a design that's part of a benchmark test suite specifically for image->html conversion. The dense visual noise / complexity / flowing svg shapes are things that most LLMs have trouble with.
It's meant to be a good test, not a good design.
It’s not really AI slop, it’s how most modern SAAS websites look like.
Seeing people talking about the Agents SDK -> credits change makes me wonder does it impact Zed or likes.
At work we use haiku 4.5 for a handful of latency sensitive tasks that are fairly simple. It performs well. Just started testing 5.5 as I’ve been anticipating a nice improvement since it was teased. Results so far are trash. Prompt leakage even. And it’s slower. I guess it’s cheap but I think they got the balance wrong on this.
The monthly API credits for Max plan seems fantastic, especially considering Haiku pricing. Being able to actually use my Claude plan for other harnesses and use-cases on top of regular CC usage is everything I wanted.
Anthropic has really been doing all the right things in the past few weeks, while OpenAI continues to fumble the bag.
Note that this is Anthropic Trojan-Horsing the previously announced June change in with a model release, where the Claude Agent SDK can no longer be used with Claude subscriptions and is now billed with API credits only.
https://support.claude.com/en/articles/15036540-use-the-clau...
Ah, that sucks. I'm using Paseo to run Claude Code; I guess that just got a whole lot more complicated.
yeah totally agree. esp how efficient it can be to have a subscription quota-paid orch spin up a bunch of API agents, this is kind of like free money to encourage what was already an easy way to save money (via batch pricing)
This is great! Been using GPT 6 Luna for decompiling my childhood favorite game (Age of Mythology) and this means I can throw Haiku into the mix as well. 17352/21965 functions matched so far...
i love Age of Mythology, but why did you feel the need to decompile it? Its got a great world editor if you were trying to "mod" it.
can you share more details? was this very involved or asking codex/claude/open code with a 1 shot like approach?
I'll write a blogpost when I actually have it working, but basically I gave the game .msi installer to claude opus 5.5 and said to read these blogs:
And to setup a harness that will decompile the game and start doing a matching decompilation of every function. It set up a bunch of tooling and started a service in the background to do this actual decompilation campaign. I put some instructions into the main opus chat now and then to e.g. add automatic git pushing including a nice svg chart of progress and to switch model strategies here and there i.e. to do a first pass with a cheap model and then switch to opus/sol if the small model can't solve it.I could now one-shot a new game, yeah.
About time Anthropic released a competitive cheap model. Haiku 4.5 has been too expensive compared to its performance for months now (in fact I don't remember being too impressed even when it was released). This one actually looks worth using in some scenarios. If it's really as much of a step up from Luna as the benchmarks they've shown indicate, it'll probably replace Luna in my workflows. 100k tokens is a pretty low threshold before the price goes up, but I tend to use these smaller models for smaller tasks anyway.
> Claude Haiku 5.5 is our fastest model to date at each model’s standard speed, although it runs less quickly than our Opus models in Fast Mode.
Opus 5.5 runs 117 tps average on Openrouter, so it must be at least 10-20 tps slower for them to mention. IDK why they mention this as it does not help for marketing though. https://openrouter.ai/anthropic/claude-opus-5.5
Maybe they think it's of interest.
131tok/s P50 according to OpenRouter currently, though might move up or down over the coming days. If it sticks at that speed, roughly twice the throughput of Luna is impressive, though the 5x price increase beyond 100k is painful.
Was a big fan of Haiku 4.5, though understand why for most Sonnet was the far better option.
I wonder if we have an AI LLM equivalent to Moore's Law. Like how often do we expect improvement in this technology and with what timing?
Yes -> every 18 months they've gotten 90% more efficient for the same level of quality for about 5 years. There's little sign that trend is slowing. If anything, there's reason to believe that System 1 models (plus potentially 1-2-3 workflows) may increase that over the next 3-5 years.
You'll know when the trend stops -> when the intelligence differential between smaller models like 7B starts to grow instead of shrink from 32B models -> that means 7B is getting about as smart as it can get. Then, 32B will follow next, then 70B, etc etc.
We haven't yet seen that at any size AFAIK.
Andrej Karpathy said once that he expects superintelligence could fit in 1 billion parameters.
Super intelligence that doesn't have to deal with the real world, maybe.
I wouldn't be surprised if less than 1B param equivalent of our brain deals with solving math and writing computer programs and physics and all the things we tend to associate with "intelligence" - especially if you ultra optimized for that, I doubt our brain works like that.
Dealing with the real world, I highly highly doubt it.
How about if we get away from written text as the input, to something more fundamental, that then also is able to produce text (among other things)?
Given that humans learn to talk while having encountered a measly number of word instances, and, given enough time, we should always be able to improve on the lottery that is biology, it does seems fairly likely.
According to Epoch AI:
> The cost of achieving a given level of AI performance has fallen about 47% per quarter since 2023, or 13× per year. [0]
0. https://epoch.ai/publications/the-plunging-price-of-thought
Then why are AI plans still so super expensive, and AI spending going through the roof, while all the subsidies are ending?
Because it's increasingly useful and the thing you are displacing (human time) is much more expensive.
The cost per fixed level of intelligence is dropping, but we're also getting dramatically more intelligent models.
At least for me the Claude plans seem like an incredible deal and I never hit my limit.
I've heard tell about 100% of certain types of work being ended in batches of six months. For years. Truthfully, I'm skeptical, but accuracy wasn't prioritized.
reminds me of this blog post: https://campedersen.com/singularity
double the information density every 2 days?
serious bit: if you think about how these smaller models work, at the end of the day it seems that they are now capable of forgetting useless information because they're able to derive it in reasoning allowing models to become smaller at the cost of requiring more reasoning tokens to solve a task.
Knowledge will be shifted to systems like n-gram augmentation which are relatively cheap and will not compete with reasoning capabilities for weight saturation.
Hopefully enough runway for an existing model to train the next to be better than itself with absolutely no human intervention.
Good to see Anthropic back alternative OSes.
Top of the page in 17 minutes? Now I know what y'all do while your agents are working.
Apparently quite a bit smarter than Luna, I wonder what use cases it can cover. I actually honestly don't need a Haiku level AI to be that smart, and looks like you pay for it in the per token cost, I need speed mainly. I might even rather have a dumber but much faster model for things like web searching and parsing to retrieve results for the app or other LLM to do things with.
From these selected benchmarks, it looks like it smokes Luna capability-wise. Excited to put it through its paces
It's finally here ! Need to take a look at some benchmark now
Where is Pelican? ehehhe
This is what AI psychosis has reduced readers to on this site.
We are witnessing the acceptance of average and accelerating more of the same low quality slop.
I think it's a joke at this point, but also the visual benchmark is a remarkably dense method for demonstrating how good a model is.
Yeah, people like to poop on the pelican. But pelican quality still correlated with overall model capabilities reasonably well, and you can immediately see and interpret it. It's a running gag, but it also does have some actual value.
So do you have the pelican or no?
Both you and the OP are asking the wrong questions. You are requesting a benchmark that tests for nothing.
Lets not focus on slop benchmarks and instead focus on more rigorous ones such as KernelBench [0], CompileBench [1], TraderBench [2] and Harvey's Legal Agent Benchmark [3] as frontier-level benchmarks that are far better to verify whilst requiring expert level intelligence as the barrier.
[0] https://kernelbench.com/
[1] https://www.compilebench.com/
[2] https://openreview.net/forum?id=JokkCjyadB
[3] https://www.harvey.ai/blog/introducing-harveys-legal-agent-b...
How are y'all using Haiku though? I rarely select it.
I've been using GPT-6 Luna in some capacity for nearly all my agent workflows. It's just a really good model, and the pricing is cheap. If Haiku 5.5 is better, and the same price (under 100k context... which is a big caveat) i'd probably swap it.
It’s absolutely better than Luna. It feels closer to a “sonnet 5.2” if that makes sense.
Of course it’s not as big, and hence falls-off quicker. I’d consider the 100k a “promotional price” to match Luna’s token pricing while delivering noticeably more intelligence.
I have a zsh functions that calls claude code with haiku to suggest commit messages, is faster and the instructions are two lines.
I also have an "ask" script that I use daily to ask simple stuff, it can access websearch and webfetch, it's more than enough to parse logs, ask for commands, quick research on the internet, small stuff. https://github.com/mariocesar/dotfiles/blob/main/common/.loc...
I use haiku for things that needs to be quick, have really clear instructions.
Opus often picks it when it's doing a "find me something" subagent. But largely it's been held back by being fully a year old at this point, and priced at a much higher price than models that are far more capable.
not haiku, but luna - last week i used it for things like "read this historical dump of 15k support tickets and break them into categories that make sense, then propose help docs that i could write to handle the most frequent queries in each category"
used <10% of my 5hr limit on a $100 codex plan.
I was waiting for this.
Planning on doing flash analyses of PRs that impact evals in some way, and then post comments on GitHub whenever there’s flaws in them
( https://evalship.com )
I'm building a game that incorporates LLMs as a game mechanic.
I've been using Luna, but I'll probably switch to Haiku.
It's great at parsing documents inexpensively. For the few skills/plugins I've made, I usually instruct Claude to use Haiku for low-reasoning grunt work.
The forgotten model is back on the map. I actually got OK mileage when I tried it for coding months ago. Maybe I'll try it again, with Opus guiding it, and see how it goes.
It fails the "How many r's in <word>?" test.
I ask:
> how many r's in diminished
It answers:
> Diminished has 1 r.
will use it to replace luna in production !
It’s about time Haiku got an update!
The important question though...how does it do making a pelican on a bicycle?
How does the price compare to Luna? At least looking at the numbers it is noticeably better at most tasks.
IMO, this is better. Luna is super cheap, but it's not that capable. At higher levels of reasoning, it's not that fast.
This is more expensive, but it also looks like it's better enough that it's far more useful.
I also won't be surprised if you look at cost per completed task + wall clock time that it comes out ahead for the majority of what you'd want to actually use it for.
Luna will still be a great option for doing non-engineering tasks super cheaply.
For prompts under 100k tokens, it's priced the same as Luna - $0.10 in, $0.50 out.
For prompts over 100k tokens it's 5 times more expensive - $0.50 in, $2.50 out.
Happy about the Sonnet cache read price cut.
That was effectively required to match GPT-6.1 Sol (costs and caching prices are now equal). Sonnet 5.5 made zero sense to use over Opus 5.5 under the old cache prices.
Wow, the rate of improvements in the AI era is staggering.
GDPval-AA v2.1 as of now: 1620
GDPval-AA v2.1 for Haiku 4.5: 735
The 100k tokens pricing makes sense, looks to be a hedge against OpenAI's decisions API and Jev or its open source alternatives that are springing up.
Nice release, congrats to Anthropic.
> Second, this week, we’ll roll out a new monthly API credit to all Max and Team subscribers for use on the Claude Platform. Max 5x users will get $100 in credits per month, Max 20x users will get $200, and Team subscribers will receive up to $500, pooled across their users. These credits are designed to allow our users to experiment with building tools, apps, and agents that call our API. They can be used on any of our models. For more information, see our Help Center article.
Did anyone read this? We get free API credits on some plans now
The important question is, does it talk in incomprehensible Claude-ese like the other Claude 5.x models?
Probably the same scam as the last Haiku update I guess. Uses more tokens to compensate for the lower price.
> Haiku 5.5’s cybersecurity safeguards are more restrictive than Haiku 4.5’s, but somewhat less restrictive than those we’ve applied to other recent models. In cybersecurity, they permit a wider range of defensive tasks than our safeguards for Sonnet 5.5, but they still block penetration testing and other techniques more likely to be used by attackers.
If you block pentest or "other techniques more likely to be used by attackers", then what does "permit a wider range of defensive tasks" even mean?
Any defensive task that's meaningful is almost indistinguishable from legitimate red-teaming that then falls under 'likely to be used by attackers". If only they would just stop nerfing these models, that'd be great. No APT is waiting around for Anthropic's permission, so might as well let us have some cool stuff.
Yeah it's too little too late, cat's out of the bag as people know that GLM 5.3 exists and is great at defensive and offensive cybersec.
(sadly Mistral Large 4 isn't up to par - but Mistral serves GLM at 130 tps!)
These are the examples from "Detecting and countering misuse of AI: September 2026" - https://news.ycombinator.com/item?id=49647300
Anthropic then says for the above, "we estimate that the uplift provided by Claude was primarily clerical assistance in data analysis, study ideation and design"While doing my best to avoid comment, please note, they're talking about a domain expert in a state research institution using Claude to do paperwork.
What did they save us from? What bioweapons did these filters prevent? From the front matter report,
OK. Sounds serious. "Gain of function research..." but who and why? So this was a researcher inside of some country's national lab ("credible institutional context") doing research on dangerous viruses using Claude for "for editorial assistance in writing up the research."What "uplift" are you providing to scientists working at specialized global BSL-4 labs that already have – and I quote their report - "physical access to such isolates." (as in samples of viruses)?
This nonsense has been expanded with even worse "safeguards."
This makes Claude unusable for any serious scientist or people curious about science, which is sad.
I remember a friend asking me why LLMs suck so bad. She was using Haiku 4.5 and that poor model couldn't keep track of the context within 3 messages.
She said she was using Haiku 4.5 because she was advised to be careful with the spending.
I hate that model so much lol.
Does anyone still use Haiku model?
I use it for title generation basically. Will have to see where this one fits in
Opus 5.5 does :^)