API Reference

Fyntune's REST API is available at https://api.fyntuneq.com/v1. All requests must include your API key in the Authorization header.

Authentication

Include your API key as a Bearer token on every request:

Authorization: Bearer ck_your_api_key_here

POST /evals/run

Submit a single LLM call for evaluation. Returns an eval run ID. Results are available within 90 seconds.

Request

{
  "feature": "summarization",
  "version": "v12",
  "input": "The quarterly report showed...",
  "output": "Q3 revenue increased 12%...",
  "model": "gpt-4o"
}

Parameters

Parameter	Type	Required	Description
feature	string	Yes	Feature name as defined in your fyntune.yaml
version	string	Yes	Prompt version tag (e.g. "v12", "prod-2026-04")
input	string	Yes	The input passed to the LLM
output	string	Yes	The LLM's response
model	string	No	Model identifier for tracking (e.g. "gpt-4o")

Response

{
  "run_id": "run_abc123",
  "status": "queued",
  "estimated_completion_ms": 72000
}

GET /evals/runs/{run_id}

Retrieve the result of a completed eval run.

Response

{
  "run_id": "run_abc123",
  "status": "complete",
  "verdict": "pass",
  "completed_at": "2026-06-20T14:32:01Z",
  "criteria": [
    {
      "name": "semantic_similarity",
      "score": 0.874,
      "threshold": 0.820,
      "status": "pass"
    }
  ]
}

GET /evals/runs

List recent eval runs for your project. Supports pagination.

Query parameters

Parameter	Type	Default	Description
feature	string	all	Filter by feature name
version	string	all	Filter by prompt version
limit	integer	20	Number of runs to return (max 100)
after	string		Cursor for pagination (run_id)

Error codes

Code	Meaning
401	Invalid or missing API key
404	Run ID not found
429	Rate limit exceeded (100 req/min per API key)
503	Eval worker temporarily unavailable — retry with exponential backoff