openrouter-model-availability

OpenRouter Model Availability

Overview

OpenRouter's /api/v1/models endpoint is the source of truth for model availability. Models can be temporarily unavailable, have degraded performance, or be permanently removed. This skill covers querying model status, building health probes, tracking availability over time, and automating failover.

Prerequisites

• An OpenRouter API key exported as OPENROUTER_API_KEY for live probes (the catalog query itself needs no auth) — see the openrouter-install-auth skill for setup
• curl and jq for the catalog status queries and the cron monitoring script
• Python 3.8+ with the OpenAI SDK and requests (pip install openai requests) for the health-check service
• A small credit balance — each max_tokens: 1 probe costs roughly $0.0001

Instructions

1. Confirm your models exist in the catalog with curl -s https://openrouter.ai/api/v1/models | jq ... per Query Model Status — pull context_length and per-million pricing without spending any tokens.
2. For a zero-cost existence check inside code, use check_model_exists() from Catalog-Based Availability Check; on a miss it calls find_similar() to suggest same-provider replacements.
3. Probe live health with probe_model() from Health Check Service — a max_tokens: 1 request that returns a HealthStatus with available, latency_ms, and checked_at.
4. Sweep your critical set with check_critical_models(), which logs OK/FAIL plus latency per model.
5. Automate via the Availability Monitoring Script as a */5 * * * * cron job appending timestamped status lines to /var/log/openrouter-health.log.
6. Tune alerting per Error Handling — require 2-3 consecutive failures before marking a model down to avoid false positives.

Query Model Status

# Check if specific models exist and their status
curl -s https://openrouter.ai/api/v1/models | jq '[.data[] | select(
  .id == "anthropic/claude-3.5-sonnet" or
  .id == "openai/gpt-4o" or
  .id == "openai/gpt-4o-mini"
) | {
  id,
  context_length,
  prompt_per_M: ((.pricing.prompt | tonumber) * 1000000),
  completion_per_M: ((.pricing.completion | tonumber) * 1000000)
}]'

# List all available models (just IDs)
curl -s https://openrouter.ai/api/v1/models | jq '[.data[].id] | sort'

# Count models by provider
curl -s https://openrouter.ai/api/v1/models | jq '[.data[].id | split("/")[0]] | group_by(.) | map({provider: .[0], count: length}) | sort_by(-.count)'

Health Check Service

import os, time, logging
from datetime import datetime, timezone
from dataclasses import dataclass
import requests
from openai import OpenAI, APIError, APITimeoutError

log = logging.getLogger("openrouter.health")

@dataclass
class HealthStatus:
    model: str
    available: bool
    latency_ms: float
    checked_at: str
    error: str = ""

client = OpenAI(
    base_url="https://openrouter.ai/api/v1",
    api_key=os.environ["OPENROUTER_API_KEY"],
    timeout=15.0,
    default_headers={"HTTP-Referer": "https://my-app.com", "X-Title": "health-check"},
)

def probe_model(model_id: str) -> HealthStatus:
    """Send a minimal request to test model availability."""
    start = time.monotonic()
    try:
        response = client.chat.completions.create(
            model=model_id,
            messages=[{"role": "user", "content": "hi"}],
            max_tokens=1,  # Minimal cost
        )
        latency = (time.monotonic() - start) * 1000
        return HealthStatus(
            model=model_id, available=True, latency_ms=round(latency, 1),
            checked_at=datetime.now(timezone.utc).isoformat(),
        )
    except (APIError, APITimeoutError) as e:
        latency = (time.monotonic() - start) * 1000
        return HealthStatus(
            model=model_id, available=False, latency_ms=round(latency, 1),
            checked_at=datetime.now(timezone.utc).isoformat(),
            error=str(e),
        )

def check_critical_models() -> list[HealthStatus]:
    """Probe all critical models."""
    CRITICAL_MODELS = [
        "anthropic/claude-3.5-sonnet",
        "openai/gpt-4o",
        "openai/gpt-4o-mini",
        "google/gemini-2.0-flash-001",
    ]
    results = []
    for model in CRITICAL_MODELS:
        status = probe_model(model)
        log.info(f"{'OK' if status.available else 'FAIL'} {model} ({status.latency_ms}ms)")
        results.append(status)
    return results

Catalog-Based Availability Check

def check_model_exists(model_id: str) -> dict:
    """Check if a model exists in the catalog (no API call cost)."""
    resp = requests.get("https://openrouter.ai/api/v1/models")
    models = {m["id"]: m for m in resp.json()["data"]}

    if model_id in models:
        m = models[model_id]
        return {
            "exists": True,
            "context_length": m["context_length"],
            "pricing": m["pricing"],
        }
    return {"exists": False, "suggestion": find_similar(model_id, models)}

def find_similar(model_id: str, models: dict) -> list[str]:
    """Find models with similar names (for migration when model is removed)."""
    prefix = model_id.split("/")[0]
    return [m for m in models if m.startswith(prefix)][:5]

Availability Monitoring Script

#!/bin/bash
# Run as cron job: */5 * * * * /path/to/check_models.sh

MODELS=("anthropic/claude-3.5-sonnet" "openai/gpt-4o" "openai/gpt-4o-mini")
LOG_FILE="/var/log/openrouter-health.log"

for MODEL in "${MODELS[@]}"; do
  START=$(date +%s%N)
  HTTP_CODE=$(curl -s -o /dev/null -w "%{http_code}" \
    https://openrouter.ai/api/v1/chat/completions \
    -H "Authorization: Bearer $OPENROUTER_API_KEY" \
    -H "Content-Type: application/json" \
    -d "{\"model\":\"$MODEL\",\"messages\":[{\"role\":\"user\",\"content\":\"ping\"}],\"max_tokens\":1}" \
    --max-time 15)
  END=$(date +%s%N)
  LATENCY=$(( (END - START) / 1000000 ))

  STATUS="OK"
  [ "$HTTP_CODE" != "200" ] && STATUS="FAIL"

  echo "$(date -u +%Y-%m-%dT%H:%M:%SZ) $STATUS $MODEL $HTTP_CODE ${LATENCY}ms" >> "$LOG_FILE"
done

Output

• HealthStatus records per probed model: available, latency_ms, an ISO-8601 checked_at timestamp, and the error string when a model is down
• Catalog check dicts: {"exists": True, "context_length": ..., "pricing": ...} on a hit, or {"exists": False, "suggestion": [...]} listing similar model IDs on a miss
• Append-only log lines from the cron script, e.g. 2026-07-02T14:05:01Z OK anthropic/claude-3.5-sonnet 200 842ms, one per critical model every 5 minutes

Examples

Check that a critical model is still in the catalog before spending tokens on a probe:

curl -s https://openrouter.ai/api/v1/models | jq '[.data[] | select(
  .id == "anthropic/claude-3.5-sonnet") | {id, context_length}]'
# [{"id": "anthropic/claude-3.5-sonnet", "context_length": 200000}]

Then run the Python health sweep — run_health_checks() in references/examples.md prints [OK] anthropic/claude-3.5-sonnet: 842.3ms per model, a 3/3 models healthy summary, and the mapped fallback (e.g. openai/gpt-4-turbo) for any failure. More worked examples: references/examples.md.

Error Handling

Enterprise Considerations

• Health probes cost tokens ($0.0001 or less per probe with max_tokens: 1) -- budget for monitoring
• Require 2-3 consecutive failures before marking a model as down to avoid false positives
• Cache the models list and refresh every 5 minutes -- don't hit /api/v1/models on every request
• Subscribe to OpenRouter announcements for model deprecations and new additions
• Maintain a model alias map so your code uses logical names (e.g., "primary-chat") that you can remap
• Alert when critical models disappear from the catalog, not just when they fail probes

References

• Examples | Errors
• Models API | Status