Provider reliability · tracked daily
DeepSeek: DeepSeek V4.1 Flash (batch): which provider is most reliable?
Across 40 providers, Together is the most reliable way to run DeepSeek: DeepSeek V4.1 Flash (batch), at 100.0% uptime and 208ms median latency. Measured directly from OpenRouter's endpoint data, tracked daily.
- Together - 100.0% uptime · 208ms
- Wafer - 100.0% uptime · 218ms
- ModelRun - 100.0% uptime · 317ms
- Modal - 100.0% uptime · 357ms
- Parasail - 100.0% uptime · 365ms
- Sail Research - 100.0% uptime · 399ms
- Makora - 100.0% uptime · 400ms
- Relace - 100.0% uptime · 415ms
- Fireworks - 100.0% uptime · 419ms
- Crusoe - 100.0% uptime · 565ms
Uptime is OpenRouter's own trailing-30-minute metric, captured once at our last daily sync - a snapshot of that window, not a live/real-time feed.
| # | Provider | Uptime (30m) | Latency p50 | Throughput |
|---|---|---|---|---|
| 01 | Together | 100.0% | 208ms | 193 tok/s |
| 02 | Wafer | 100.0% | 218ms | 185 tok/s |
| 03 | ModelRun | 100.0% | 317ms | 137 tok/s |
| 04 | Modal | 100.0% | 357ms | 163 tok/s |
| 05 | Parasail | 100.0% | 365ms | 169 tok/s |
| 06 | Sail Research | 100.0% | 399ms | 96 tok/s |
| 07 | Makora | 100.0% | 400ms | 110 tok/s |
| 08 | Relace | 100.0% | 415ms | 97 tok/s |
| 09 | Fireworks | 100.0% | 419ms | 173 tok/s |
| 10 | Crusoe | 100.0% | 565ms | 191 tok/s |
| 11 | CoreWeave | 100.0% | 591ms | 168 tok/s |
| 12 | OpenInference | 100.0% | 603ms | 103 tok/s |
| 13 | Cohere | 100.0% | 610ms | 167 tok/s |
| 14 | Alibaba | 100.0% | 624ms | 116 tok/s |
| 15 | Morph | 100.0% | 648ms | 84 tok/s |
| 16 | DigitalOcean | 100.0% | 672ms | 176 tok/s |
| 17 | Krea | 100.0% | 710ms | 96 tok/s |
| 18 | DeepSeek | 100.0% | 734ms | 61 tok/s |
| 19 | Cloudflare | 100.0% | 736ms | 116 tok/s |
| 20 | AtlasCloud | 100.0% | 741ms | 89 tok/s |
| 21 | Ionstream | 100.0% | 832ms | 124 tok/s |
| 22 | Mancer 2 | 100.0% | 839ms | 79 tok/s |
| 23 | PrimeIntellect | 100.0% | 839ms | 148 tok/s |
| 24 | InferenceNet | 100.0% | 890ms | 74 tok/s |
| 25 | Nebius | 100.0% | 945ms | 139 tok/s |
| 26 | Novita | 100.0% | 1061ms | 115 tok/s |
| 27 | Baidu | 100.0% | 1160ms | 86 tok/s |
| 28 | Reka | 100.0% | 1306ms | 91 tok/s |
| 29 | 100.0% | 1368ms | 75 tok/s | |
| 30 | StreamLake | 100.0% | 1490ms | 62 tok/s |
| 31 | DekaLLM | 100.0% | 1567ms | 95 tok/s |
| 32 | SiliconFlow | 100.0% | 1634ms | 85 tok/s |
| 33 | GMICloud | 100.0% | 1651ms | 89 tok/s |
| 34 | Io Net | 99.5% | 1131ms | 91 tok/s |
| 35 | NextBit | 99.5% | 2640ms | 67 tok/s |
| 36 | BaseTen | 90.0% | 165ms | 167 tok/s |
| 37 | Decart | 90.0% | 711ms | 140 tok/s |
| 38 | Venice | 86.7% | 1300ms | 155 tok/s |
| 39 | DeepInfra | 0.0% | - | - |
| 40 | Phala | - | 1489ms | 75 tok/s |
Uptime over time
- Alibaba
- AtlasCloud
- Baidu
- BaseTen
- Cloudflare
- Cohere
Frequently asked questions
Which provider is most reliable for DeepSeek: DeepSeek V4.1 Flash (batch)?
Together is the most reliable provider for DeepSeek: DeepSeek V4.1 Flash (batch), at 100.0% uptime (OpenRouter's trailing-30-minute metric, snapshotted at our last daily sync) and 208ms median latency, out of 40 providers tracked.
What is DeepSeek: DeepSeek V4.1 Flash (batch)'s uptime?
The best-performing provider for DeepSeek: DeepSeek V4.1 Flash (batch) reports 100.0% uptime, tracked daily.