How long does the integration take?

Change one base URL in an OpenAI-compatible client, or install the Python or TypeScript SDK; the quickstart routes a first request in a few minutes.

Will this lock us into another vendor?

No. The router is Apache-2.0 open source, your provider keys stay yours, and the API is OpenAI-compatible, so leaving is a change of base URL.

What happens when a provider fails?

The router falls back to the next model your policy allows and records the fallback in the trace, so an outage costs a retry, not an incident.

90 days of the enterprise, pro and team plans free, no card needed. See the plans

OpenSmartRoute

Open-weights models - Template - OpenSmartRoute

Templatev1.0.0

Open-weights models

Name: Open-weights models
Author: community

Hosted open-weights models only - the same stack runs on your own hardware later.

by community(0) 0 installs

Free

Free account. Installing gives you the manifest plus copy-paste snippets.

See reviews

About

Generated from the reference model catalogue (the OpenRouter listing) - regenerated daily as the line-up and prices change. Import it with registry://<slug>@<version> or osr stack apply; every target answers on https://openrouter.ai/api/v1 with your own key.

Open-weights models

Hosted open-weights models only - the same stack runs on your own hardware later.

Targets

MiMo-V2.6-Pro (Xiaomi) - $0.43 in / $0.87 out per 1M tokens, intelligence 46
GLM 5.3 (Z.ai) - $0.08 in / $1.40 out per 1M tokens, intelligence 45
Step 5 Preview (StepFun) - $1.00 in / $2.70 out per 1M tokens, intelligence 44
Kimi K3 (Moonshot AI) - $0.24 in / $10.80 out per 1M tokens, intelligence 44

1 routing rule(s); every target answers on https://openrouter.ai/api/v1 (OpenAI-compatible).

Use it

Copy one of these into your project. Installing also returns the manifest and these snippets.

bash

osr stack plan registry://stack-open-weights@1.0.0 --registry https://osr-api.kindrock-caaab1a2.centralindia.azurecontainerapps.io
osr stack apply registry://stack-open-weights@1.0.0 --registry https://osr-api.kindrock-caaab1a2.centralindia.azurecontainerapps.io

Installed into a catalogue, chosen by a router

Install Open-weights models and it becomes one more candidate the router can pick - when it fits.

A listing is a routing target with a manifest: what it does, which domains it covers, what it costs and who publishes it. Once installed it sits beside your own models and tools, is scored like any other candidate for each request, and shows up in the trace when it wins. Ratings come from workspaces that installed it, one per account.

Photo: RDNE Stock project

For the business

What it costs to adopt - in engineering time.

Same client, same code
One base URL
The OpenAI-compatible endpoint answers your existing SDK calls; model=auto lets the router choose.
The SDKs
One URL, every vendor
No vendor lock-in
Point your OpenAI-compatible client at one address; add, swap or retire providers in the catalogue without touching application code.
Every vendor side by side
Next eligible model answers
Failover built in
When a provider is down or over quota the request moves on; the trace says it did.

stack-open-weights.stack.jsonjson

{
  "osr": "1",
  "kind": "stack",
  "name": "Open-weights models",
  "description": "Hosted open-weights models only - the same stack runs on your own hardware later.",
  "targets": [
    {
      "ocm": "1",
      "id": "model-xiaomi-mimo-v2-6-pro",
      "kind": "llm",
      "name": "MiMo-V2.6-Pro",
      "description": "MiMo-V2.6-Pro is the flagship foundation model developed by Xiaomi. Built at a scale of over 1T parameters, it is designed to push the ceiling of capability for the most demanding...",
      "publisher": "Xiaomi",
      "version": "1.0.0",
      "capabilities": {
        "domains": [
          "coding",
          "math"
        ],
        "tags": [
          "llm",
          "xiaomi",
          "openrouter",
          "open-weights",
          "reasoning",
          "tool-calling",
          "image",
          "video",
          "audio",
          "models"
        ],
        "languages": [
          "en"
        ],
        "modalities": [
          "text",
          "image",
          "video",
          "audio"
        ],
        "supports_tools": true,
        "supports_streaming": true,
        "context_window": 1050000
      },
      "quality_prior": 0.463,
      "examples": [
        "MiMo-V2.6-Pro is the flagship foundation model developed by Xiaomi. Built at a scale of over 1T parameters, it is designed to push the ceiling of capability for the most demanding..."
      ],
      "primary": false,
      "metadata": {
        "source": {
          "provider": "models",
          "ref": "1790021259",
          "url": "https://openrouter.ai/xiaomi/mimo-v2.6-pro",
          "key": "xiaomi/mimo-v2.6-pro",
          "catalogue": "https://openrouter.ai/api/v1"
        },
        "model": "xiaomi/mimo-v2.6-pro",
        "pricing": {
          "input_usd_per_1m": 0.435,
          "output_usd_per_1m": 0.87,
          "cache_read_usd_per_1m": 0.0036
        },
        "max_output_tokens": 131072,
        "output_modalities": [
          "text"
        ],
        "hugging_face_id": "XiaomiMiMo/MiMo-V2.6-Pro-RL",
        "reasoning": true,
        "created": 1790021259,
        "benchmarks": {
          "intelligence_index": 46.3,
          "hub_trending": 34,
          "hub_downloads": 96140,
          "hub_downloads_total": 105069,
          "hub_likes": 770,
          "osr_index": 47.2,
          "osr_coverage": 0.4
        }
      },
      "cost": {
        "usd_per_1k_tokens": 0.000652
      },
      "endpoints": [
        {
          "protocol": "openai",
          "url": "https://openrouter.ai/api/v1"
        }
      ]
    },
    {
      "ocm": "1",
      "id": "model-z-ai-glm-5-3",
      "kind": "llm",
      "name": "GLM 5.3",
      "description": "GLM-5.3 is a large-scale reasoning model from Z.ai, built for complex software engineering and long-horizon agent tasks. It supports text input and output with a 1M-token context window, and improves...",
      "publisher": "Z.ai",
      "version": "1.0.0",
      "capabilities": {
        "domains": [
          "coding",
          "math"
        ],
        "tags": [
          "llm",
          "z-ai",
          "openrouter",
          "open-weights",
          "reasoning",
          "tool-calling",
          "models"
        ],
        "languages": [
          "en"
        ],
        "modalities": [
          "text"
        ],
        "supports_tools": true,
        "supports_streaming": true,
        "context_window": 1048576
      },
      "quality_prior": 0.448,
      "examples": [
        "GLM-5.3 is a large-scale reasoning model from Z.ai, built for complex software engineering and long-horizon agent tasks. It supports text input and output with a 1M-token context window, and improves..."
      ],
      "primary": false,
      "metadata": {
        "source": {
          "provider": "models",
          "ref": "1787086655",
          "url": "https://openrouter.ai/z-ai/glm-5.3",
          "key": "z-ai/glm-5.3",
          "catalogue": "https://openrouter.ai/api/v1"
        },
        "model": "z-ai/glm-5.3",
        "pricing": {
          "input_usd_per_1m": 0.079,
          "output_usd_per_1m": 1.4,
          "cache_read_usd_per_1m": 0.078
        },
        "max_output_tokens": 943718,
        "output_modalities": [
          "text"
        ],
        "hugging_face_id": "zai-org/GLM-5.3",
        "reasoning": true,
        "created": 1787086655,
        "benchmarks": {
          "intelligence_index": 44.8,
          "coding_index": 74.8,
          "agentic_index": 53.1,
          "hub_likes": 2303,
          "hub_downloads": 2043783,
          "hub_downloads_total": 3203582,
          "hub_trending": 51,
          "osr_index": 46.4,
          "osr_coverage": 0.4
        }
      },
      "cost": {
        "usd_per_1k_tokens": 0.000739
      },
      "endpoints": [
        {
          "protocol": "openai",
          "url": "https://openrouter.ai/api/v1"
        }
      ]
    },
    {
      "ocm": "1",
      "id": "model-stepfun-step-5-preview",
      "kind": "llm",
      "name": "Step 5 Preview",
      "description": "Step 5 Preview is StepFun's flagship model for agentic work, built on a sparse Mixture-of-Experts architecture (27B active / 600B total parameters). It performs strongly in software engineering and professional...",
      "publisher": "StepFun",
      "version": "1.0.0",
      "capabilities": {
        "domains": [
          "coding",
          "math"
        ],
        "tags": [
          "llm",
          "stepfun",
          "openrouter",
          "open-weights",
          "reasoning",
          "tool-calling",
          "image",
          "video",
          "models"
        ],
        "languages": [
          "en"
        ],
        "modalities": [
          "text",
          "image",
          "video"
        ],
        "supports_tools": true,
        "supports_streaming": true,
        "context_window": 1048576
      },
      "quality_prior": 0.437,
      "examples": [
        "Step 5 Preview is StepFun's flagship model for agentic work, built on a sparse Mixture-of-Experts architecture (27B active / 600B total parameters). It performs strongly in software engineering and professional..."
      ],
      "primary": false,
      "metadata": {
        "source": {
          "provider": "models",
          "ref": "1791462849",
          "url": "https://openrouter.ai/stepfun/step-5-preview",
          "key": "stepfun/step-5-preview",
          "catalogue": "https://openrouter.ai/api/v1"
        },
        "model": "stepfun/step-5-preview",
        "pricing": {
          "input_usd_per_1m": 1,
          "output_usd_per_1m": 2.7,
          "cache_read_usd_per_1m": 0.1
        },
        "max_output_tokens": 64000,
        "output_modalities": [
          "text"
        ],
        "hugging_face_id": "stepfun-ai/Step-5-Preview",
        "reasoning": true,
        "created": 1791462849,
        "benchmarks": {
          "intelligence_index": 43.7,
          "hub_downloads": 2467,
          "hub_downloads_total": 2485,
          "hub_likes": 127,
          "hub_trending": 34,
          "osr_index": 45.8,
          "osr_coverage": 0.4
        }
      },
      "cost": {
        "usd_per_1k_tokens": 0.00185
      },
      "endpoints": [
        {
          "protocol": "openai",
          "url": "https://openrouter.ai/api/v1"
        }
      ]
    },
    {
      "ocm": "1",
      "id": "model-moonshotai-kimi-k3",
      "kind": "llm",
      "name": "Kimi K3",
      "description": "Kimi K3 is a 2.8T parameter open-weight multimodal reasoning model from Moonshot AI. It is suited for complex coding, knowledge work, and long-horizon agentic workflows, and is particularly strong at...",
      "publisher": "Moonshot AI",
      "version": "1.0.0",
      "capabilities": {
        "domains": [
          "coding",
          "math"
        ],
        "tags": [
          "llm",
          "moonshotai",
          "openrouter",
          "open-weights",
          "reasoning",
          "tool-calling",
          "image",
          "video",
          "models"
        ],
        "languages": [
          "en"
        ],
        "modalities": [
          "text",
          "image",
          "video"
        ],
        "supports_tools": true,
        "supports_streaming": true,
        "context_window": 1048576
      },
      "quality_prior": 0.436,
      "examples": [
        "Kimi K3 is a 2.8T parameter open-weight multimodal reasoning model from Moonshot AI. It is suited for complex coding, knowledge work, and long-horizon agentic workflows, and is particularly strong at..."
      ],
      "primary": false,
      "metadata": {
        "source": {
          "provider": "models",
          "ref": "1784215858",
          "url": "https://openrouter.ai/moonshotai/kimi-k3",
          "key": "moonshotai/kimi-k3",
          "catalogue": "https://openrouter.ai/api/v1"
        },
        "model": "moonshotai/kimi-k3",
        "pricing": {
          "input_usd_per_1m": 0.239,
          "output_usd_per_1m": 10.8,
          "cache_read_usd_per_1m": 0.21
        },
        "max_output_tokens": 943718,
        "output_modalities": [
          "text"
        ],
        "hugging_face_id": "moonshotai/Kimi-K3",
        "reasoning": true,
        "created": 1784215858,
        "benchmarks": {
          "intelligence_index": 43.6,
          "coding_index": 76.2,
          "agentic_index": 50,
          "hub_downloads": 909762,
          "hub_downloads_total": 5708037,
          "hub_likes": 11816,
          "hub_trending": 49,
          "osr_index": 45.7,
          "osr_coverage": 0.4
        }
      },
      "cost": {
        "usd_per_1k_tokens": 0.00552
      },
      "endpoints": [
        {
          "protocol": "openai",
          "url": "https://openrouter.ai/api/v1"
        }
      ]
    }
  ],
  "rules": [
    {
      "name": "hard-to-strongest",
      "when": {
        "min_complexity": 0.7
      },
      "prefer": [
        "model-xiaomi-mimo-v2-6-pro"
      ]
    }
  ],
  "objective": {
    "quality": 0.7,
    "cost": 0.8,
    "latency": 0.2
  },
  "metadata": {
    "source": {
      "provider": "catalogue-stacks",
      "path": "open-weights",
      "ref": "2026-10-24",
      "url": "https://openrouter.ai/models",
      "key": "catalogue-stacks/open-weights",
      "catalogue": "https://openrouter.ai/api/v1",
      "generated": true
    }
  }
}

Fetch it by URL: GET /api/v1/registry/stack-open-weights/manifest?version=1.0.0

Open-weights models

About

Open-weights models

Targets

Use it

Install Open-weights models and it becomes one more candidate the router can pick - when it fits.

What it costs to adopt - in engineering time.

One base URL

No vendor lock-in

Failover built in

Manifest

Reviews