Claude Fable 5.2 vs Gemini 4 Hydrogen
Claude Fable 5.2 and Gemini 4 Hydrogen are within three points on the Models at Work leaderboard, which we treat as a tie (71 vs 69 out of 100). On independent tests alone, Claude Fable 5.2 is ahead (100 vs 78). Gemini 4 Hydrogen is cheaper: 1,000 support replies cost about $7.50 at list price, against $40 on Claude Fable 5.2, which is 5.3 times as much. On the Artificial Analysis Intelligence Index, one of the independent tests, Claude Fable 5.2 leads 65 to 51.
| Claude Fable 5.2 · Anthropic | Gemini 4 Hydrogen · Google | |
|---|---|---|
| Leaderboard rank | #3 | #9 |
| Score (0–100) | 71 | 69 |
| Kind of work | Hardest problems | Everyday default |
| Price per 1M tokens | $8 to read · $40 to write | $1.50 to read · $7.50 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 | Gemini Enterprise Agent Platform (Google Cloud), Google AI Studio, Gemini for business, Gemini Enterprise, Gemini app |
| Can a company host it itself | No | No |
| Artificial Analysis Intelligence Index | 65 | 51 |
| Humanity's Last Exam, Diamond (Scale) | 72.40 | Not measured |
| Humanity's Sixth Sense (Scale) | 59.50 | Not measured |
| Epoch Capabilities Index | 173 | Not measured |
| Vals Finance Index | Not measured | 66.14% (rank 2) |
| Vals Tax Agent Bench | Not measured | 52.30% (rank 1) |
| Vals Legal Research Bench | Not measured | 53.85% (rank 6) |
| Best for | Financial and technical analysis with no room for error, Multi-step research, Work Claude Opus 5.5 gets wrong | Everyday drafting and summarising, Spreadsheet models and forecasts in Google Sheets, First-pass tax and finance research, checked by a person, Multi-step tasks across Gmail and Drive |
Cost of 1,000 jobs
| 1,000 of these | Claude Fable 5.2 | Gemini 4 Hydrogen |
|---|---|---|
| Support repliesthe customer message, the account notes and two help articles read; a 300-word reply written | $40 | $7.50 |
| Documents processeda two-page invoice or form read; the fields pulled out | $28 | $5.25 |
| Calls summariseda 40-minute call transcript read; a summary and next steps written | $104 | $20 |
| Contracts revieweda 40-page contract read against a checklist; a memo of issues written | $300 | $56 |
| Drafts writtena brief and one example read; a 600-word email or proposal section written | $44 | $8.25 |
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 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.
Choose Gemini 4 Hydrogen for
- Everyday drafting and summarising
- Spreadsheet models and forecasts in Google Sheets
- First-pass tax and finance research, checked by a person
- Multi-step tasks across Gmail and Drive
Scores 51 on Artificial Analysis's index, level with the dearer Gemini 4 Argon (50), at $1.50 in and $7.50 out per million tokens: 25% under Argon and Carbon today and 62.5% under them from 1 December.
Caveats: Released 12 November 2026: one independent test counts toward its score so far. Vals AI's tax, legal and finance tests are recorded below but not yet scored, because fewer than three models on this board carry them. Google sells it as the successor to Gemini 3.8 Flash, but it costs twice what 3.8 Flash does until 1 January 2027, when Google's pricing page says 3.8 Flash rises to the same $1.50 in / $7.50 out per million tokens. On Vals AI's tax test it ranks first with 52.3% of questions right, so close to half its answers there were wrong: a person still checks the output. In Gemini Enterprise it is switched off until an administrator turns it on, per Google Cloud's release notes. Google does not say which Gemini app plans get it or whether it becomes the default in Gemini for business.
What people running them say
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.
Gemini 4 Hydrogen: On the Hacker News (a developer forum) launch thread, 534 upvotes, most comments are about Google's model names. On substance: modeless reads Google's charts as 'a better choice than Sonnet in areas outside of coding'; onlyrealcuzzo calls early results 'disappointing', reporting it does more than asked; ComputerGuru notes it costs more than anyone pays for 3.8 Flash today. Developers on day one, none verified, and too little to score.
Questions people ask
Which is better, Claude Fable 5.2 or Gemini 4 Hydrogen?
Claude Fable 5.2 and Gemini 4 Hydrogen are within three points on the Models at Work leaderboard, which we treat as a tie (71 vs 69 out of 100). On independent tests alone, Claude Fable 5.2 is ahead (100 vs 78). Gemini 4 Hydrogen is cheaper: 1,000 support replies cost about $7.50 at list price, against $40 on Claude Fable 5.2, which is 5.3 times as much. On the Artificial Analysis Intelligence Index, one of the independent tests, Claude Fable 5.2 leads 65 to 51.
When should I choose Claude Fable 5.2 over Gemini 4 Hydrogen?
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.
When should I choose Gemini 4 Hydrogen over Claude Fable 5.2?
Choose Gemini 4 Hydrogen for everyday drafting and summarising, spreadsheet models and forecasts in google sheets, first-pass tax and finance research, checked by a person, multi-step tasks across gmail and drive. Scores 51 on Artificial Analysis's index, level with the dearer Gemini 4 Argon (50), at $1.50 in and $7.50 out per million tokens: 25% under Argon and Carbon today and 62.5% under them from 1 December.
Which is cheaper, Claude Fable 5.2 or Gemini 4 Hydrogen?
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 13, 2026). A token is about three-quarters of a word. Gemini 4 Hydrogen costs $1.50 per million tokens it reads (input) and $7.50 per million tokens it writes (output) on Google's list price (Google: Introducing Gemini 4 Hydrogen, read Nov 13, 2026). A token is about three-quarters of a word.
What do 1,000 jobs cost on Claude Fable 5.2 and Gemini 4 Hydrogen?
At list price: 1,000 support replies cost $40 on Claude Fable 5.2 and $7.50 on Gemini 4 Hydrogen; 1,000 documents processed cost $28 on Claude Fable 5.2 and $5.25 on Gemini 4 Hydrogen; 1,000 calls summarised cost $104 on Claude Fable 5.2 and $20 on Gemini 4 Hydrogen; 1,000 contracts reviewed cost $300 on Claude Fable 5.2 and $56 on Gemini 4 Hydrogen; 1,000 drafts written cost $44 on Claude Fable 5.2 and $8.25 on Gemini 4 Hydrogen. 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.