Quality
Known Good Index and the capability indices, side by side
Each index is min-max normalised across tracked models and scaled 0–100 from published evaluations. We do not run these evaluations.
| Metric | claude-sonnet-4-5-20250929 | DeepSeek: DeepSeek V4 Pro |
|---|---|---|
| Known Good Index | 59 | 93 |
| Coding Index | 61 | — |
| Mathematics Index | 89 | 97 |
| Science Index | 84 | 94 |
| Reasoning Index | 48 | — |
| Agentic Index | 61 | 89 |
| Openness Index | 5 | 60 |
Quality Breakdown
Every evaluation both models report a published score for
Bar = claude-sonnet-4-5-20250929 · marker = DeepSeek: DeepSeek V4 Pro · scores ingested from Epoch AI (CC-BY 4.0)
Right-hand figures read claude-sonnet-4-5-20250929 / DeepSeek: DeepSeek V4 Pro. claude-sonnet-4-5-20250929 leads on 0 of 3, DeepSeek: DeepSeek V4 Pro on 3.
Arena Elo
Human preference rating from blind pairwise votes
Elo from blind pairwise votes, ingested from Arena. Higher is better.
| Metric | claude-sonnet-4-5-20250929 | DeepSeek: DeepSeek V4 Pro |
|---|---|---|
| Text Arena | 1,438 | 1,449 |
| Text Arena (style controlled) | 1,455 | 1,457 |
Price & Cost
Published API pricing by token type
USD per 1M tokens. Blended is a 3:1 input:output mix. Live from OpenRouter, cross-checked against models.dev. Lower is better.
| Metric | claude-sonnet-4-5-20250929 | DeepSeek: DeepSeek V4 Pro |
|---|---|---|
| Input $/1M | — | $0.43 |
| Output $/1M | — | $0.87 |
| Blended $/1M | — | $0.54 |
| Cache read $/1M | — | $0.004 |
Context Window
Maximum input accepted and maximum output emitted
Provider-declared limits, from metadata via models.dev and OpenRouter. Higher is better.
| Metric | claude-sonnet-4-5-20250929 | DeepSeek: DeepSeek V4 Pro |
|---|---|---|
| Context window | — | 1.05M |
| Max output tokens | — | 384k |
Speed & Latency
Median throughput and time to first token across hosting providers
Medians across all tracked providers, derived from OpenRouter endpoint telemetry.
| Metric | claude-sonnet-4-5-20250929 | DeepSeek: DeepSeek V4 Pro |
|---|---|---|
| Output tokens/s | — | 45 |
| Time to first token | — | 2.0s |
Providers
Which hosts serve each model, and at what price and speed
| Provider | claude-sonnet-4-5-20250929 blended | claude-sonnet-4-5-20250929 tok/s | DeepSeek: DeepSeek V4 Pro blended | DeepSeek: DeepSeek V4 Pro tok/s |
|---|---|---|---|---|
| BaseTen | — | — | $2.17 | 110 |
| Fireworks | — | — | $2.17 | 66 |
| Novita | — | — | $1.46 | 66 |
| DeepInfra | — | — | $1.62 | 65 |
| GMICloud | — | — | $0.85 | 62 |
| Venice | — | — | $2.06 | 61 |
| Baidu · first party | — | — | $0.78 | 58 |
| Together | — | — | $2.17 | 53 |
| Alibaba · first party | — | — | $1.77 | 52 |
| AtlasCloud | — | — | $2.10 | 46 |
| Wafer | — | — | $1.50 | 45 |
| DeepSeek · first party | — | — | $0.54 | 44 |
| SiliconFlow | — | — | $1.91 | 41 |
| StreamLake | — | — | $0.84 | 41 |
| CoreWeave | — | — | $2.17 | 28 |
| Ionstream | — | — | $1.41 | 20 |
| DigitalOcean | — | — | $1.74 | 6 |
| Parasail | — | — | $2.17 | — |
Per-host pricing and telemetry from OpenRouter. A dash means that host does not serve that model.