DeepSeek: DeepSeek V4.1 Flash (batch)
deepseek/deepseek-v4-1-flash-batchDeepSeek V4.1 Flash is a sparse mixture-of-experts model from DeepSeek, and the first built on the company's Causal Encoder-Decoder (CED) architecture.
Modalities
In / out per 1M$0.05 / $0.15
Context1M
ReleasedSep 10, 2026
Providers
39 providers serve this model upstream.
| Provider | Input /M | Output /M | Cache read /M | Latency P50 | Throughput P50 | Uptime |
|---|---|---|---|---|---|---|
| $0.04 | $0.50 | $0.01 | 0.32s | 188 tps | 100% | |
| $0.19 | $0.65 | $0.01 | 0.23s | 196 tps | 100% | |
 DigitalOcean | $0.23 | $0.90 | $0.01 | 0.69s | 175 tps | 100% |
| $0.30 | $1.20 | $0.01 | 0.22s | 211 tps | 100% | |
 DeepInfraus · fp8 | $0.14 | $0.42 | $0.00 | 0.61s | 102 tps | 100% |
 Cohere | $0.30 | $1.20 | $0.00 | 0.22s | 167 tps | 100% |
| $0.30 | $1.20 | $0.01 | 0.45s | 181 tps | 100% | |
| $0.10 | $0.60 | $0.01 | 0.31s | 100 tps | 100% | |
 Baseten | $0.30 | $1.20 | $0.01 | 0.37s | 181 tps | 100% |
| $0.20 | $0.78 | $0.00 | 0.93s | 108 tps | 100% | |
| $0.10 | $0.60 | $0.01 | 0.33s | 78 tps | 100% | |
| $0.15 | $0.59 | $0.00 | 1.60s | 75 tps | 100% | |
| $0.14 | $0.54 | $0.01 | 0.67s | 64 tps | 100% | |
| $0.30 | $1.20 | $0.03 | 0.46s | 125 tps | 100% | |
| $0.09 | $0.25 | $0.01 | — | — | 100% | |
| $0.09 | $0.25 | $0.01 | — | — | 100% | |
 Fireworksus | $0.45 | $1.80 | $0.01 | 0.64s | 237 tps | 100% |
| $0.30 | $1.20 | $0.01 | 0.86s | 164 tps | 99.97% | |
| $0.20 | $1.00 | Variable | 1.17s | 50 tps | 99.96% | |
| $0.05 | $0.19 | $0.01 | 0.59s | 91 tps | 99.95% | |
| $0.29 | $1.20 | $0.01 | 0.47s | 125 tps | 99.95% | |
 SiliconFlowfp8 | $0.30 | $1.20 | $0.01 | 1.46s | 81 tps | 99.92% |
| $0.00 | $0.76 | $0.00 | 2.49s | 56 tps | 99.87% | |
| Fireworks Spot | $0.21 | $0.84 | $0.00 | 0.61s | 132 tps | 99.8% |
| $0.30 | $1.20 | $0.01 | 0.26s | 139 tps | 99.8% | |
| DeepInfraus · fp8 | $0.11 | $0.34 | $0.00 | 2.42s | 54 tps | 99.78% |
| Basetenus | $0.30 | $1.20 | $0.01 | 0.34s | 172 tps | 99.76% |
| $0.10 | $0.84 | $0.01 | 0.30s | 91 tps | 99.65% | |
 Google Vertexglobal | $0.30 | $1.20 | $0.03 | 0.76s | 60 tps | 99.54% |
| $0.26 | $1.17 | $0.01 | 0.25s | 129 tps | 99.49% | |
| $0.00 | $0.76 | $0.00 | 0.67s | 115 tps | 99.17% | |
| $0.30 | $1.20 | $0.01 | 0.82s | 94 tps | 99.14% | |
| $0.17 | $0.66 | $0.02 | 0.64s | 131 tps | 99.08% | |
| $0.30 | $1.20 | $0.01 | 0.79s | 173 tps | 98.99% | |
 CoreWeavefp8 | $0.13 | $0.65 | $0.00 | 0.56s | 205 tps | 97.83% |
| Fireworks | $0.30 | $1.20 | $0.01 | 0.62s | 132 tps | 97.2% |
| $0.21 | $0.84 | $0.00 | 1.07s | 70 tps | 95% | |
| $0.18 | $0.72 | $0.00 | 2.52s | 86 tps | 94.14% | |
| DeepInfraus · fp8 | $0.21 | $0.63 | $0.01 | 0.50s | 86 tps | 74.26% |
Market-wide measurements, refreshed every 15 minutes. They describe who serves DeepSeek V4.1 Flash (batch) in general, not how ProjectCOZY routes your requests.
Pricing
List rates per 1M tokens — what DeepSeek V4.1 Flash (batch) costs pay-as-you-go elsewhere, not what you pay here.
Input
$0.05
per 1M tokens
Output
$0.15
per 1M tokens
Cache read
$0.00
per 1M cached tokens
Included in every pass
DeepSeek V4.1 Flash (batch) is included in every ProjectCOZY pass at no extra cost — one flat rate, no per-token metering. The rates above are there for comparison.
Performance
Time to first token and output speed across every provider, over the last 3 days.
0.0s20.0s40.0s60.0s80.0s
Oct 29, 12 AMOct 29, 8 AMOct 29, 4 PMOct 30, 12 AMOct 30, 8 AMOct 30, 4 PMOct 31, 12 AMOct 31, 8 AMOct 31, 4 PMNov 1, 12 AMNov 1, 8 AMNov 1, 4 PM
inference.net (us) Fireworks Spot Makora Wafer (us) NovitaAI Sail Research (us) Cohere Mancer SiliconFlow Baidu Qianfan Alibaba Cloud Int. Wafer Crusoe Phala Prime Intellect AtlasCloud Baseten Morph (us) StreamLake Google Vertex (global) DeepInfra (us) Baseten (us) Relace Open Inference Venice Parasail Fireworks (us) Together DeepInfra (us) DeepInfra (us) Modal CoreWeave GMICloud Sail Research DigitalOcean Reka AI Nebius Token Factory inference.net Fireworks
0 tok/s2000 tok/s4000 tok/s6000 tok/s8000 tok/s
Oct 29, 12 AMOct 29, 8 AMOct 29, 4 PMOct 30, 12 AMOct 30, 8 AMOct 30, 4 PMOct 31, 12 AMOct 31, 8 AMOct 31, 4 PMNov 1, 12 AMNov 1, 8 AMNov 1, 4 PM
inference.net (us) Fireworks Spot Makora Wafer (us) NovitaAI Sail Research (us) Cohere Mancer SiliconFlow Baidu Qianfan Alibaba Cloud Int. Wafer Crusoe Phala Prime Intellect AtlasCloud Baseten Morph (us) StreamLake Google Vertex (global) DeepInfra (us) Baseten (us) Relace Open Inference Venice Parasail Fireworks (us) Together DeepInfra (us) DeepInfra (us) Modal CoreWeave GMICloud Sail Research DigitalOcean Reka AI Nebius Token Factory inference.net Fireworks
Uptime
100% of requests were served successfully over the last 3 days, across all providers.
Last 3 daysAvailability 99.67%
Oct 29, 6:00 PMNow
Availability over the last 24 hoursAvailability 99.23%
100% 97% 93% 90%
Oct 31, 6:10 PMNow
Apps
Tools that most often run DeepSeek V4.1 Flash (batch), by tokens this week.
Activity
117069.6B tokens processed market-wide in the last month.
02000.0B4000.0B6000.0B8000.0B
Oct 2Oct 7Oct 12Oct 17Oct 22Oct 27Nov 1
Tokens
FAQ
Which model id do I use for DeepSeek V4.1 Flash (batch)?
deepseek/deepseek-v4.1-flash:batch. Send it as the model field from any OpenAI-compatible client, against https://minuykh.app/v1.
How much context does DeepSeek V4.1 Flash (batch) hold?
1,048,576 tokens per request, of which up to 384,000 can be the response.
What can it accept and return?
Input: text, image. Output: text.
What does DeepSeek V4.1 Flash (batch) cost on a pass?
Nothing beyond the pass itself. Pay-as-you-go elsewhere it lists at $0.05 per 1M input tokens and $0.15 per 1M output tokens.
Who built DeepSeek V4.1 Flash (batch)?
DeepSeek, in the DeepSeek family.
Explore
Other models your pass can call.
