Claude Fable 5 vs DeepSeek V3.2
Head to head on 20 shared benchmarks, with price, context window, and release dates.
Benchmark scores
53evals · sort by any column, ⤢ to expand fullscreen
| Model | ||||||||||||||||||||||||||||||||||||||||||||||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Claude Fable 5 Anthropic | 71.8% | 90.0% | 1933elo | 83 | 92.6% | 55.5% | 63.5% | 81.4% | 78.6% | 80.0% | 59.9% | 88.5% | 93.9% | 87.3% | 95.0% | 60.2% | 76.9% | 62.9% | 90.4% | 98.5% | - | - | - | 17.4% | 83.9% | 38.6% | - | 1340elo | 78.0% | 56.3% | 29.3% | - | 29.8% | 1932elo | 66.0% | - | - | 13.3% | - | - | 56.1% | 88.5% | - | 68.9% | 85.0% | 51.9% | - | - | 80.3% | 88.0% | - | 66.0% | 74.2% | 69.4% |
| DeepSeek V3.2 DeepSeek | 47.9% | 16.0% | 1511elo | 66.8 | 75.1% | 11.2% | 49.0% | 65.9% | 64.6% | 50.0% | 48.2% | 70.4% | 85.0% | 77.2% | 8.0% | 38.7% | 68.2% | 32.6% | 5.1% | 78.9% | 74.7% | 74.2% | 59.0% | - | - | - | 0.0% | - | - | - | - | 22.1% | - | - | - | 31.3% | 68.0% | - | 59.3% | 57.5% | - | - | 83.7% | - | - | - | 59.0% | 70.0% | - | - | 65 | - | - | 51.7% |
2 / 2 models
| Attribute | Claude Fable 5 Anthropic | DeepSeek V3.2 DeepSeek |
|---|---|---|
| Description | Claude Fable 5 is Anthropic's most capable widely released model, built for the most demanding reasoning and long-horizon agentic work. Shares its base model with the invitation-only Claude Mythos 5. | DeepSeek V3.2 is an AI model from DeepSeek, released with open weights. |
| Family | Claude | Deepseek |
| Params | - | - |
| Open weights | closed | open |
| License | - | mit |
| Released | 9 Jun 2026 | 1 Dec 2025 |
| Context | 1,000,000 | 163,840 |
| Input $/1M | $10.00 | $0.28 |
| Output $/1M | $50.00 | $0.42 |
| Cache read $/1M | $1.000 | $0.130 |
| Cache write $/1M | $12.500 | - |
| Speed | 66 tok/s | 0 tok/s |
| Latency | 49.62 s | 0.00 s |
| Modalities | text, image, file | text |
| Providers | - | - |
| Publisher | Anthropic | DeepSeek |
| Reported scores | Finance Agent (v2)56.3% IFBench63.5% (AA) Swebenchpro80.3% LiveBench - Math93.9% (LiveBench) LiveBench - Coding78.6% (LiveBench) | Aider Polyglot Benchmark74.2% (Aider) MathArena57.47% GPQA Diamond75.1% (AA) Humanity's Last Exam (HLE)11.2% (AA) |
FAQ
- Is Claude Fable 5 or DeepSeek V3.2 better?
- Across the 20 benchmarks both models report on Sophon, Claude Fable 5 leads on 20 of them. Which one is "better" depends on the benchmark - the table above shows every shared score.
- How do Claude Fable 5 and DeepSeek V3.2 compare on GPQA Diamond?
- Claude Fable 5 scores 92.6% and DeepSeek V3.2 scores 75.1% on GPQA Diamond.
- Which is cheaper, Claude Fable 5 or DeepSeek V3.2?
- DeepSeek V3.2 is cheaper at $0.42 per million output tokens against $50.00 for Claude Fable 5 - about 119.0x.
- When were Claude Fable 5 and DeepSeek V3.2 released?
- Claude Fable 5 was released 9 Jun 2026 by Anthropic. DeepSeek V3.2 was released 1 Dec 2025 by DeepSeek.
Comparing something else? Build your own side-by-side.