DeepSeek V4.1 Flash model for reasoning and agentic coding
Benchmarks
| Benchmark | Category | Score | Variant | Provenance | Source | Observed |
|---|---|---|---|---|---|---|
| AA Agentic Index | agentic | 41.0 | — | mirror | openrouter | Sep 20, 2026 |
| AA Coding Index | coding | 69.1 | — | mirror | openrouter | Jul 31, 2026 |
| AA Intelligence Index | intelligence | 34.3 | — | mirror | openrouter | Sep 20, 2026 |
| Design Arena — 3d | preference | 1237 | — | mirror | openrouter | Sep 12, 2026 |
| Design Arena — asciiart | preference | 1111 | — | mirror | openrouter | Sep 17, 2026 |
| Design Arena — codecategories | preference | 1244 | — | mirror | openrouter | Sep 21, 2026 |
| Design Arena — dataviz | preference | 1196 | — | mirror | openrouter | Sep 20, 2026 |
| Design Arena — gamedev | preference | 1233 | — | mirror | openrouter | Sep 18, 2026 |
| Design Arena — svg | preference | 1212 | — | mirror | openrouter | Sep 20, 2026 |
| Design Arena — uicomponent | preference | 1253 | — | mirror | openrouter | Sep 20, 2026 |
| Design Arena — website | preference | 1251 | — | mirror | openrouter | Sep 16, 2026 |
| LMArena Text | preference | 1423 | thinking | crowd | lmarena | Jul 30, 2026 |
| 1432 | — | crowd | lmarena | Sep 13, 2026 | ||
| LMArena WebDev | preference | 1580 | high-effort | crowd | lmarena | Sep 11, 2026 |
Includes index data from Artificial Analysis.
Providers
Live operational stats per hosting endpoint, via OpenRouter.
| Provider | Quant | Context | $/1M in · out | Uptime 24h | Latency | Throughput |
|---|---|---|---|---|---|---|
| Alibaba | — | 1M | $0.352 · $1.06 | 99.96% | — | — |
| Alibaba | fp8 | 1M | $0.134 · $0.268 | 98.95% | — | — |
| AtlasCloud | fp4 | 1.0M | $0.14 · $0.28 | 99.96% | — | — |
| Azure | — | 1.0M | $0.21 · $0.56 | 97.80% | — | — |
| Baidu | fp8 | 1.0M | $0.14 · $0.28 | 99.92% | — | — |
| BaseTen | fp8 | 1.0M | $0.13 · $0.26 | 100.00% | — | — |
| Cloudflare | — | 1.3M | $0.44 · $1.32 | 99.99% | — | — |
| CoreWeave | fp8 | 262K | $0.13 · $0.28 | 99.99% | — | — |
| DeepInfra | fp8 | 1.0M | $0.09 · $0.18 | 99.51% | — | — |
| DigitalOcean | — | 1.0M | $0.098 · $0.196 | 99.94% | — | — |
| Fireworks | — | 1.0M | $0.22 · $0.66 | 97.79% | — | — |
| GMICloud | fp8 | 1.0M | $0.091 · $0.182 | 99.87% | — | — |
| Inceptron | fp4 | 1.0M | $0.08 · $0.2 | 99.40% | — | — |
| Makora | — | 1M | $0.09 · $0.195 | 98.80% | — | — |
| Mancer 2 | fp8 | 1.0M | $0.19 · $0.5 | 96.86% | — | — |
| Morph | — | 1.0M | $0.142 · $0.4 | 93.50% | — | — |
| NextBit | fp8 | 1.0M | $0.15 · $0.35 | 98.25% | — | — |
| Novita | fp8 | 1.0M | $0.14 · $0.28 | 99.98% | — | — |
| OpenInference | fp8 | 1.0M | $0.06 · $0.14 | 99.73% | — | — |
| Parasail | fp8 | 1.0M | $0.14 · $0.28 | 99.70% | — | — |
| Phala | — | 1.0M | $0.44 · $1.32 | 99.91% | — | — |
| Reka | fp4 | 262K | $0.088 · $0.528 | 99.59% | — | — |
| Relace | fp4 | 1.0M | $0.04 · $0.16 | 99.70% | — | — |
| Sail Research | fp4 | 1.0M | $0.074 · $0.342 | 99.60% | — | — |
| SiliconFlow | fp8 | 1.0M | $0.13 · $0.28 | 97.89% | — | — |
| StreamLake | fp8 | 1.0M | $0.089 · $0.177 | 96.96% | — | — |
| Together | — | 1.0M | $0.14 · $0.28 | 99.75% | — | — |
| Venice | — | 1M | $0.097 · $0.193 | 96.74% | — | — |
| Wafer | — | 1.0M | $0.1 · $0.25 | 99.97% | — | — |