Mistral: Mixtral 8x22B Instruct
mistralai/mixtral-8x22b-instruct- Modalities
- text+file->text
- In / out price
- $2 / $6
- Context
- 66K
- Released
- Apr 17, 2024
- Knowledge cutoff
- 2024-01-31
per 1M
Mistral's official instruct fine-tuned version of Mixtral 8x22B. It uses 39B active parameters out of 141B, offering unparalleled cost efficiency for its size. Its strengths include: - strong math, coding,...
Supported parameters
frequency_penaltyFrequency penalty
Lowers the chance of a token in proportion to how often it already appeared.
max_tokensMax tokens
Upper bound on generated tokens. Output stops at the limit.
presence_penaltyPresence penalty
Lowers the chance of any token that has already appeared.
response_formatResponse format
For example json_object or json_schema, to constrain the shape.
seedSeed
A fixed seed helps reproduce sampling. It does not guarantee byte-identical output.
stopStop
Stop when these strings are produced. A string or an array of strings.
structured_outputsStructured outputs
Generate a parseable structure that follows a JSON Schema.
temperatureTemperature
Sampling temperature. Lower is steadier, higher is more varied. Often 0 to 2.Default 0.3
tool_choiceTool choice
Whether to call a tool: auto, none, or a named function.
toolsTools
Function definitions the model may call via tool_calls.
top_pTop-p
Sample only from tokens whose cumulative probability reaches top_p.