Lab vs Lab
OpenAI vs Anthropic.
The club-level face-off: Club Scores, best models, squad depth, division titles and the price floor — every number computed from the model database.
| OpenAI | Anthropic | Edge | |
|---|---|---|---|
| Club Score | 94.4 | 95.8 | Anthropic |
| Best model | GPT-6 Astra | Claude Fable 5.1 | Anthropic |
| Top-3 average | 90.9 | 93.5 | Anthropic |
| Squad size | 6 | 4 | OpenAI |
| Cheapest capable ($/1M blended) | 0.5 | 0.2 | Anthropic |
| Open-weight share % | 0 | 0 | Draw |
| Division titles | 1 | 1 | Draw |
OpenAI holds the #1 model in: Reasoning Division.
Anthropic holds the #1 model in: Value Division.
Price floor — cheapest capable model on three workloads
| Workload | OpenAI | Anthropic |
|---|---|---|
| Chat assistant (1M in / 250K out) | $0.50 | $0.23 |
| Document analysis (5M in / 500K out) | $1.60 | $0.75 |
| Code generation (2M in / 1000K out) | $1.60 | $0.70 |
Best vs best
League fixture: The two clubs' #1 models, head-to-head. In Season Zero, fixture results are decided by the data — the category winners above. When voted fixtures launch, results blend 60% benchmark / 40% crowd per the published rulebook.
Verdict by category
| Category | GPT-6 Astra | Claude Fable 5.1 | Winner |
|---|---|---|---|
| League Score | 96.7 | 97.3 | Claude Fable 5.1 |
| Coding (SWE-bench %) | — | — | — |
| Reasoning (GPQA %) | 96.0 | 93.7 | GPT-6 Astra |
| Context (tokens) | 1050000 | 1000000 | GPT-6 Astra |
| Speed (tok/s) | — | — | — |
| Price (blended $/1M) | $20.00 | $20.00 | Claude Fable 5.1 (cheaper) |
Tale of the tape
| Spec | GPT-6 Astra | Claude Fable 5.1 |
|---|---|---|
| Maker | OpenAI | Anthropic |
| Released | Sep 3, 2026 | Sep 1, 2026 |
| League Score | 97 | 97 |
| Context | 1,050,000 tokens | 1,000,000 tokens |
| Max output | 128,000 | 128,000 |
| $/1M input | $10.00 | $10.00 |
| $/1M output | $50.00 | $50.00 |
| Weights | Closed | Closed |
| Speed | — | — |
Head-to-head benchmarks (shared only)
| Benchmark | GPT-6 Astra | Claude Fable 5.1 |
|---|---|---|
| LiveBench overall | 86.1 | 88.0 |
| GPQA Diamond | 96.0 | 93.7 |
Cost at three workloads
| Workload | GPT-6 Astra | Claude Fable 5.1 |
|---|---|---|
| Light — 1M in / 250K out | $22.50 | $22.50 |
| Standard — 10M in / 2.5M out | $225.00 | $225.00 |
| Long-context — 50M in / 5M out | $750.00 | $750.00 |
The pick
Pick GPT-6 Astra if you need the bigger context window.
Pick Claude Fable 5.1 if you want the higher overall League Score.
All figures from the model database — sources and dates on each model page. Cost your own workload or compare a different pair.