deepseek-v4.1-flash-1017
deepseek-v4.1-flash-1017Availability100.0%
Available
Samples: 16
Latency1.2 s
Output speed128.8 tok/s
DeepSeek V4.1 Flash (2026-10-17 snapshot): sparse MoE flash tier with 1M context. Reasoning model: it always emits reasoning before the answer, so allow generous max_tokens.
- Input / 1M tokens
- $0.44
- Output / 1M tokens
- $1.32
- Context window
- 1,048,576
- Route
- Default
DefaultRoleplay
Not measuredNo matching public benchmark is available for this model.
Routes
Ability score divided by token cost, ranked across measured model routes. Cost uses equal input and output token counts; higher is better value.
| Use this model | ||||||
|---|---|---|---|---|---|---|
| devin | $0.22 | $0.66 | $0.007 | Not measured | Not measured | Plans and tiers |
| forge | $0.22 | $0.66 | $0.007 | Not measured | Not measured | Plans and tiers |
| ollama_pro | $0.22 | $0.66 | $0.007 | Not measured | Not measured | Plans and tiers |
| Default | $0.44 | $1.32 | $0.014 | Not measured | Not measured | Use this model |
devin
ollama_proQuick start
Access and billingcurl --fail-with-body 'https://api.tokenreply.com/v1/chat/completions' \
-H "Authorization: Bearer $TOKENREPLY_API_KEY" \
-H 'Content-Type: application/json' \
--data '{
"model": "deepseek-v4.1-flash-1017",
"messages": [
{
"role": "user",
"content": "Reply with hello."
}
]
}'