Claude Fable 5
Timed at 54.6 tok/s, scored 76.5 on the Coding Index, on 18 Design Arena boards, and with results on 2 openly licensed boards.
retrieved sha256 · 0bb53a77713f
The same model sold by 2 providers, each at its own list price. OpenRouter's rate is its routed rate across whichever of them it uses, and it need not match any one of them.
Measured speed
368 requests · 5 endpoints · 30-minute windowOpenRouter's own timing of its routed traffic, weighted by requests, over one 30-minute window read . Every route is on the endpoint ledger below.
Every endpoint
5 priced · 5 timed · cheapest first- Anthropicanthropic · listed$10.00$50.00$1.00not declared128k45 tok/s30.23.05s73 req100%
- Azureazure · listed$10.00$50.00$1.00not declared128k48 tok/s26.32.44s1 reqno reading
- Claude Platform on AWSclaude-on-aws · listed$10.00$50.00$1.00not declared128k57 tok/s374.56s286 req100%
- Googlegoogle-vertex/global · listed$10.00$50.00$1.00not declared128k60 tok/s45.11.14s5 req100%
- Googlegoogle-vertex/global · listed$10.00$50.00$1.00not declared128k50 tok/s385.42s3 req100%
Each track runs from a route's p90, the rate nine in ten of its requests met or beat, up to its median. Weighted by requests, the figure above reports where the traffic went rather than what one route does: 3 of the 5 timed routes are slower than it, and the middle route runs at 50 tok/s. The widest median-to-p90 gap is Azure's, 1.83x.
Uptime is OpenRouter's own figure for each seller over the trailing day, read once a day, and the column prints the latest reading, taken . 2 of the 6 sellers with a reading fell under 99% on at least one of the 23 days held, 23 Oct 2026 to 21 Nov 2026: Anthropic at 98.64% on 13 Nov 2026 and Claude Platform on AWS at 98.95% on 13 Nov 2026. Source: OpenRouter's public endpoints API, one reading a day for each seller: 1,211 sellers carried one on the latest of the 23 days held.
Source: OpenRouter's public endpoints API. 1257 sellers across 359 models when the mirror ran, 186 of them served by more than one. List prices, not what anybody paid; a discount is the seller's own. Read . Speed is OpenRouter's own timing over one 30-minute window, read .
What each seller accepts
11 parameters · 4 sellers- include_reasoningevery seller
- max_tokensevery seller
- reasoningevery seller
- reasoning_effortevery seller
- response_formatevery seller
- stopevery seller
- structured_outputsevery seller
- tool_choiceevery seller
- toolsevery seller
- verbosityevery seller
- max_completion_tokens1 of 4not Anthropic, Claude Platform on AWS, and Google
10 of these 11 are accepted by every one of the 4 sellers that publish a list; the other 1 are declared by some sellers and not others, so what a request sending one of them gets depends on which seller serves it. Each list is the seller's own declaration in OpenRouter's endpoints API, read , and nothing here tested one.
Price position
342 priced modelsBlended $20.00 per million tokens, 3 in to 1 out: cheaper than 7 of the 342 priced models in the catalog. Each bar is one stretch of the catalog on a logarithmic price scale; the mark is this listing.
Measured quality
145 models with an index · 132 on a design arena boardEach mark is one board this model stands on, placed at the share of that board it scores above, and led to its number in the list under the rule. The columns and the ledger below carry the figures.
- Agentic game development · Agents1231 Elowin rate 62.0% · source rank 2the board, 25 ranked
- Agentic HTML slides · Agents1254 Elowin rate 59.4% · source rank 1the board, 9 ranked
- Agentic slides, HTML · Agents1252 Elowin rate 59.5% · source rank 1the board, 9 ranked
- Android native · Agents1276 Elowin rate 65.3% · source rank 4the board, 41 ranked
- Full stack · Agents1254 Elowin rate 58.2% · source rank 9the board, 44 ranked
- Godot game development · Agents1261 Elowin rate 64.3% · source rank 2the board, 34 ranked
- HTML slides · Agents1174 Elowin rate 63.2% · source rank 4the board, 25 ranked
- Mobile apps · Agents1254 Elowin rate 60.4% · source rank 8the board, 49 ranked
- python-pptx slides · Agents1269 Elowin rate 61.2% · source rank 4the board, 29 ranked
- Web apps · Agents1242 Elowin rate 56.3% · source rank 14the board, 47 ranked
- 3D scenes · Models1297 Elowin rate 61.6% · source rank 19the board, 113 ranked
- ASCII art · Models1310 Elowin rate 67.9% · source rank 6the board, 67 ranked
- Code · Models1300 Elowin rate 57.9% · source rank 16the board, 120 ranked
- Data visualisation · Models1303 Elowin rate 57.8% · source rank 12the board, 110 ranked
- Game development · Models1324 Elowin rate 60.8% · source rank 14the board, 120 ranked
- SVG · Models1292 Elowin rate 62.8% · source rank 10the board, 83 ranked
- UI components · Models1308 Elowin rate 58.2% · source rank 16the board, 116 ranked
- Websites · Models1287 Elowin rate 56.3% · source rank 22the board, 126 ranked
Position 13 of the 258 models Epoch has placed, which is not the population of either board above, on a scale anchored at Claude 3.5 Sonnet 130 and GPT-5 150. Epoch states that absolute values are meaningless by themselves and that comparisons between models are what the index is for, that values are not linearly related to benchmark accuracy, and that a highly specialised model may score low while being capable in its own domain.
Benchmark results
2 of 6 boards · 44 results| Task | All runs | Effort | Score |
|---|---|---|---|
| Chess Puzzles | 167 | max | 0.47 |
| high | 0.45 | ||
| low | 0.31 | ||
| EBR-bench | 34 | max | 0.40 |
| FrontierMath-2025-02-28-Public | 87 | max | 1 |
| FrontierMath-Erdos | 16 | max | 0 |
| FrontierMath-Tier-4-2025-07-01-Public | 48 | max | 1 |
| FrontierMath-Tier-4-v2-Private | 84 | max | 0.90 |
| FrontierMath-Tiers-1-3-v2-Private | 128 | max | 0.87 |
| Furniture Assembly | 42 | max | 0.36 |
| GPQA diamond | 342 | max | 0.86 |
| high | 0.83 | ||
| low | 0.79 | ||
| MirrorCode | 48 | max | 0.70 |
| high | 0.64 | ||
| Mystery Game Puzzles | 149 | max | 0.52 |
| OEIS Open Lite | 5 | high | 0.44 |
| OTIS Mock AIME 2024-2025 | 319 | max | 1.00 |
| high | 1 | ||
| low | 0.98 | ||
| SimpleQA Verified | 106 | xhigh | 0.71 |
Mean accuracy, 0 to 1, on the task the row names, as Epoch AI runs it and averaged over repeated runs. All eighteen of their tasks, including the held-out FrontierMath tiers, named as they name them. The model id carries the effort setting the run used. Epoch AI Benchmarking Hub names these runs claude-fable-5_max, claude-fable-5_xhigh, claude-fable-5_high and claude-fable-5_low, and the effort beside each score says which run it is. A run missing from a task's rows has no published result there, which is not a score of zero. All runs is how many runs the publisher scored on the task, every model's. Read from epoch.ai ↗, which dates its own data . Epoch AI Benchmarking Hub data, by Epoch AI, under CC BY 4.0.
| Task | Score |
|---|---|
| AMPS_Hard | 99 |
| code_completion | 80.44 |
| code_generation | 91.55 |
| connections | 99.33 |
| consecutive_events | 91.36 |
| integrals_with_game | 97 |
| javascript | 68.18 |
| logic_with_navigation | 78 |
| math_comp | 95.10 |
| olympiad | 92.84 |
| paraphrase | 77.65 |
| plot_unscrambling | 78.72 |
| python | 65 |
| simplify | 72.03 |
| spatial | 96 |
| story_generation | 71.70 |
| summarize | 81.70 |
| tablejoin | 56.13 |
| tablereformat | 94.12 |
| theory_of_mind | 84.61 |
| typescript | 53.33 |
| typos | 94 |
| zebra_puzzle | 100 |
Percent correct on the task the row names, higher is better. All 23 published columns; there is no average to carry, because the site computes its headline in the browser from a weighting this does not reimplement. LiveBench names this run claude-fable-5-max-effort. Read from livebench.ai ↗, which dates its own data . LiveBench, Apache-2.0. Results produced by the LiveBench maintainers on questions refreshed monthly.
2 of the 6 openly licensed boards Vitna mirrors carry a result for Claude Fable 5, each printed as its publisher published it and none added to another. MTEB, HELM, Vectara hallucination leaderboard and Aider polyglot have not published one under a name that matches its listing, and names are matched to the catalog exactly and never guessed. A task that is a link opens its column on /capability, with every run scored on it; an effort that is a link opens its run's row there, where the run is ranked on Vitna's scale. Every board, with its licence and what its publisher measures, is on /benchmarks.
Source: OpenRouter (openrouter.ai/models), as of 17 Nov 16:29 UTC. sha256 0bb53a77713f over 853,270 bytes.
Catalog: OpenRouter's public model list.
Intelligence, coding and agentic indexes: Artificial Analysis, as published by OpenRouter.
Elo, win rate and rank: Design Arena, as published by OpenRouter.
Source: the Epoch AI Capabilities Index (epoch.ai/eci), as of 20 Nov 2026. 258 models placed over 2917 benchmark results.
Epoch AI Benchmarking Hub data, by Epoch AI, under CC BY 4.0.
Vitna ran none of these evaluations and adjusted none of these figures.
