DeepSeek V4.1 Flash
Abliterated API
OpenAI- and Anthropic-compatible API with streaming, tool calls and 1M-token context.
Our lowest effective rates
| Input | Cached input | Output |
|---|---|---|
| $0.08 | $0.01 | $0.32 |
Ultra: $500/month for $1,250 credit. Use all credit before monthly renewal for these rates.
Pay as you go instead
Add funds from $0.50. Per 1M tokens: $0.20 input, $0.025 cached, $0.80 output. Balance does not expire while your account is open.
Model details
We host DeepSeek V4.1 Flash Abliterated, and it works with your existing API client.
- Model ID
deepseek-v4.1-flash-abliterated- Context / maximum output
- 1,048,576 / 131,072 tokens
- Request formats
- OpenAI Chat Completions, Responses and Anthropic Messages
- Rate limit
- 600 requests per minute per key
This is an uncensored DeepSeek API: hosted access, not GPU rental. Read what abliteration changes. The Acceptable Use Policy applies. Output pricing includes reasoning tokens.
Make your first request
Sign in and create an API key. Every new account gets $5 API credit once, with no card required. Keep your key in a private environment variable on your server.
curl --fail-with-body https://api.excessionlabs.ai/v1/chat/completions \
-H "Authorization: Bearer $EXCESSION_API_KEY" \
-H "Content-Type: application/json" \
-d '{
"model": "deepseek-v4.1-flash-abliterated",
"messages": [{"role": "user", "content": "Explain HTTP 401."}],
"max_tokens": 2048
}'For SDK examples and integration details, read the DeepSeek API guide or API reference.