Claude Haiku 5.5 vs Claude Fable 5.2
Claude Haiku 5.5 and Claude Fable 5.2 are within three points on the Models at Work leaderboard, which we treat as a tie (73 vs 71 out of 100). On independent tests alone, Claude Fable 5.2 is ahead (100 vs 55). Claude Haiku 5.5 is cheaper: 1,000 support replies cost about $0.50 at list price, against $40 on Claude Fable 5.2, which is 80 times as much. On the Artificial Analysis Intelligence Index, one of the independent tests, Claude Fable 5.2 leads 65 to 36.
| Claude Haiku 5.5 · Anthropic | Claude Fable 5.2 · Anthropic | |
|---|---|---|
| Leaderboard rank | #2 | #3 |
| Score (0–100) | 73 | 71 |
| Kind of work | High-volume work | Hardest problems |
| Price per 1M tokens | $0.10 to read · $0.50 to write | $8 to read · $40 to write |
| Reads at once | 1M tokens (about 750,000 words) | 1M tokens (about 750,000 words) |
| Where you can get it | Claude Team and Enterprise, Amazon Bedrock, Gemini Enterprise Agent Platform (Google Cloud), Microsoft Foundry, GitHub Copilot | Claude Team and Enterprise, Amazon Bedrock, Gemini Enterprise Agent Platform (Google Cloud), Microsoft Foundry |
| Can a company host it itself | No | No |
| Artificial Analysis Intelligence Index | 36 | 65 |
| Humanity's Last Exam, Diamond (Scale) | Not measured | 72.40 |
| Humanity's Sixth Sense (Scale) | Not measured | 59.50 |
| Epoch Capabilities Index | Not measured | 173 |
| Best for | Sorting and routing tickets, Pulling fields out of invoices, First-pass support replies, Tagging and classifying records at volume | Financial and technical analysis with no room for error, Multi-step research, Work Claude Opus 5.5 gets wrong |
Cost of 1,000 jobs
| 1,000 of these | Claude Haiku 5.5 | Claude Fable 5.2 |
|---|---|---|
| Support repliesthe customer message, the account notes and two help articles read; a 300-word reply written | $0.50 | $40 |
| Documents processeda two-page invoice or form read; the fields pulled out | $0.35 | $28 |
| Calls summariseda 40-minute call transcript read; a summary and next steps written | $1.30 | $104 |
| Contracts revieweda 40-page contract read against a checklist; a memo of issues written | $3.75 | $300 |
| Drafts writtena brief and one example read; a 600-word email or proposal section written | $0.55 | $44 |
Worked out from each vendor’s list price and a fixed size for each job. A product built on a model charges its own price on top. See what every model costs for the same jobs.
Choose Claude Haiku 5.5 for
- Sorting and routing tickets
- Pulling fields out of invoices
- First-pass support replies
- Tagging and classifying records at volume
The pick for high-volume routine work: a twentieth of Claude Sonnet 5.5's list price, with users reporting near-perfect sorting in production, as long as each request stays short.
Caveats: The price is five times higher on any request that feeds it more than 100,000 tokens of text at once (a long document or a long back-and-forth). It scores 36 on Artificial Analysis's index against 55 for Sonnet 5.5: keep it off work that needs judgement. It ranks 2nd because price is 30% of the board's score and this is one of the two cheapest models here, not because of its test results.
Choose Claude Fable 5.2 for
- Financial and technical analysis with no room for error
- Multi-step research
- Work Claude Opus 5.5 gets wrong
The strongest model on the board for problems the others get wrong: first on every independent test read, at twice the list price of Claude Opus 5.5.
Caveats: Anthropic's own advice is to start with Opus 5.5 and move up only when its answers fall short. It writes more per answer: Artificial Analysis's test run cost $26.54 per task at the top effort setting against $10.04 for Opus 5.5, a wider gap than the list prices suggest. On Claude Team it is offered on premium seats, up to half the weekly usage limit; Enterprise bills usage at API rates, per Anthropic's pricing page. Mythos 5.2 is the same model without the cyber and biology safeguards, open only through Anthropic's verification programs. It ranks third, not first, because price is 30% of the score and this is the second most expensive model here; on independent test results alone it is first.
What people running them say
Claude Haiku 5.5: A Hacker News (developer forum) thread on it passed a thousand upvotes in a day. The recurring reports are that it sorts items into fixed categories almost without error in live use and is 'noticeably smarter than GPT-6 Luna'; the recurring complaint is the higher price on requests over 100,000 tokens (the unit vendors bill by).
Claude Fable 5.2: One Hacker News (developer forum) launch thread, read through Signals: a commenter who tried it says it pulls ahead of Opus 5.5 only on very ambitious requests, another says Opus 5.5 at a high effort setting was enough for a software project of 125,000 lines, and a third asks how anyone measures the difference at the price. Too little to score.
Questions people ask
Which is better, Claude Haiku 5.5 or Claude Fable 5.2?
Claude Haiku 5.5 and Claude Fable 5.2 are within three points on the Models at Work leaderboard, which we treat as a tie (73 vs 71 out of 100). On independent tests alone, Claude Fable 5.2 is ahead (100 vs 55). Claude Haiku 5.5 is cheaper: 1,000 support replies cost about $0.50 at list price, against $40 on Claude Fable 5.2, which is 80 times as much. On the Artificial Analysis Intelligence Index, one of the independent tests, Claude Fable 5.2 leads 65 to 36.
When should I choose Claude Haiku 5.5 over Claude Fable 5.2?
Choose Claude Haiku 5.5 for sorting and routing tickets, pulling fields out of invoices, first-pass support replies, tagging and classifying records at volume. The pick for high-volume routine work: a twentieth of Claude Sonnet 5.5's list price, with users reporting near-perfect sorting in production, as long as each request stays short.
When should I choose Claude Fable 5.2 over Claude Haiku 5.5?
Choose Claude Fable 5.2 for financial and technical analysis with no room for error, multi-step research, work claude opus 5.5 gets wrong. The strongest model on the board for problems the others get wrong: first on every independent test read, at twice the list price of Claude Opus 5.5.
Which is cheaper, Claude Haiku 5.5 or Claude Fable 5.2?
Claude Haiku 5.5 costs $0.10 per million tokens it reads (input) and $0.50 per million tokens it writes (output) on Anthropic's list price (Claude pricing, read Nov 12, 2026). A token is about three-quarters of a word. Claude Fable 5.2 costs $8 per million tokens it reads (input) and $40 per million tokens it writes (output) on Anthropic's list price (Claude pricing, read Nov 12, 2026). A token is about three-quarters of a word.
What do 1,000 jobs cost on Claude Haiku 5.5 and Claude Fable 5.2?
At list price: 1,000 support replies cost $0.50 on Claude Haiku 5.5 and $40 on Claude Fable 5.2; 1,000 documents processed cost $0.35 on Claude Haiku 5.5 and $28 on Claude Fable 5.2; 1,000 calls summarised cost $1.30 on Claude Haiku 5.5 and $104 on Claude Fable 5.2; 1,000 contracts reviewed cost $3.75 on Claude Haiku 5.5 and $300 on Claude Fable 5.2; 1,000 drafts written cost $0.55 on Claude Haiku 5.5 and $44 on Claude Fable 5.2. A product built on either model charges its own price on top.
Scores come from the Models at Work leaderboard: Score is 0 to 100 and computed, not typed. 40% is independent testing (Artificial Analysis, Arena, Scale, Epoch, Vals and others; each test is scaled against the best score on this board, and tests of real work, such as finished freelance projects or professional tasks, count double the exams). 30% is price: the vendor's list price for a typical mix of reading and writing, on a fixed scale from $0.05 to $60 per million tokens, so a model that costs a fifth as much gains real ground. 15% is what people report after running it, from Signals and linked threads. 15% is fit: how much text it can take in at once and where a business can get it (the products and clouds that carry it), with a small credit for open weights. A missing component drops out and the rest are reweighted. A model with no independent test yet is provisional and ranks below every measured one. Within three points is a tie. Vendor-published figures are labelled and never counted. Prices are vendor list prices where published; a typical mix assumes three parts reading to one part writing.