Quality
How DeepSeek: DeepSeek V4 Pro compares against every tracked model
Composite of GPQA Diamond, Mock AIME 2024–25, MATH Level 5, SWE-bench Verified, LiveBench, Humanity's Last Exam and Terminal-Bench, normalised 0–100. Scores are ingested from Epoch AI; we do not run these evaluations.
Quality Breakdown
DeepSeek: DeepSeek V4 Pro on each evaluation feeding the index, against the tracked average
Bar = DeepSeek: DeepSeek V4 Pro · grey marker = average across tracked models · scores ingested from Epoch AI
Arena Elo
Human preference rating from blind pairwise votes
Elo from blind pairwise votes, ingested from Arena
Ingested 29 Jul 2026.
Price & Cost
Published API pricing by token type
USD per 1M tokens at a 3:1 input:output blend. Live from OpenRouter, cross-checked against models.dev. Lower is better.
Context Window
Maximum input tokens accepted
Maximum input context from provider metadata via models.dev and OpenRouter. Higher is better.
Speed
Median output throughput across hosting providers
Median output tokens per second across all tracked providers. Derived from OpenRouter endpoint telemetry. Higher is better.
Latency
Time to first token, median across providers
Seconds to first streamed chunk, median across tracked providers. Derived from OpenRouter endpoint telemetry. Lower is better.
Providers
Hosts serving DeepSeek: DeepSeek V4 Pro, with their own price and speed
| Provider | Input $/M | Output $/M | Blended | Speed | TTFT | Context |
|---|---|---|---|---|---|---|
| BaseTen | $1.74 | $3.48 | $2.17 | 110 | 0.7s | 262k |
| Parasail | $1.74 | $3.48 | $2.17 | 65 | 0.9s | 1.05M |
| Cloudflare | $1.74 | $3.48 | $2.17 | 64 | 0.7s | 393k |
| Novita | $1.17 | $2.34 | $1.46 | 62 | 1.6s | 1.05M |
| Venice | $1.65 | $3.30 | $2.06 | 55 | 1.3s | 1.00M |
| Fireworks | $1.74 | $3.48 | $2.17 | 53 | 1.9s | 1.05M |
| DeepInfra | $1.30 | $2.60 | $1.62 | 51 | 0.8s | 1.05M |
| Alibaba · first party | $1.61 | $3.22 | $2.01 | 50 | 1.5s | 1.00M |
| GMICloud | $0.68 | $1.36 | $0.85 | 48 | 2.5s | 1.05M |
| Together | $1.74 | $3.48 | $2.17 | 45 | 0.6s | 512k |
| SiliconFlow | $1.50 | $3.13 | $1.91 | 41 | 1.6s | 1.05M |
| CoreWeave | $1.74 | $3.48 | $2.17 | 40 | 0.8s | 1.05M |
| DeepSeek · first party | $0.43 | $0.87 | $0.54 | 40 | 1.1s | 1.05M |
| Baidu · first party | $0.63 | $1.25 | $0.78 | 38 | 1.1s | 1.05M |
| StreamLake | $0.67 | $1.34 | $0.84 | 38 | 1.9s | 1.02M |
| Wafer | $1.20 | $2.40 | $1.50 | 28 | 5.1s | 1.05M |
| AtlasCloud | $1.68 | $3.38 | $2.10 | 25 | 2.0s | 1.05M |
| DigitalOcean | $1.39 | $2.78 | $1.74 | 8 | 2.4s | 262k |
| Ionstream | $1.13 | $2.26 | $1.41 | 5 | 2.1s | 1.05M |
Per-host pricing and telemetry from OpenRouter. Speed is output tokens per second; TTFT is time to first token.
History
How price and quality have moved since we began tracking this model
Since 25 Jul 2026
| Known Good Index | 93 · based on 3 of 7 evaluations |
| Blended price | $0.54 per 1M · observed 29 Jul 2026 |
| Output speed | 44 tok/s · observed 29 Jul 2026 |
| Time to first token | 1.9s |
No recorded changes for this model since tracking began.
Comparisons
Head-to-head pages pairing DeepSeek: DeepSeek V4 Pro with other tracked models