ORBModels
Every model. One gateway.
18 curated models from 7 providers, addressable by id through one OpenAI-compatible endpoint — or let orbit/auto choose.
Catalog
Prices, context and capabilities.
| Capabilities | Availability | |||||
|---|---|---|---|---|---|---|
| OpenAI | gpt-oss-120b (open weights) openai/gpt-oss-120b · tier 1 | 131K | $0.024 | $0.089 | Tools | Unknown |
| DeepSeek | DeepSeek V4.1 Flash deepseek/deepseek-v4.1-flash · tier 1 | 1.05M | $0.045 | $0.50 | Tools Vision | Unknown |
| Anthropic | Claude Haiku 5.5 anthropic/claude-haiku-5.5 · tier 2 | 1M | $0.10 | $0.50 | Tools Vision | Unknown |
| Meta | Llama 4 Scout meta-llama/llama-4-scout · tier 1 | 1.31M | $0.10 | $0.30 | Tools Vision | Unknown |
| Meta | Llama 4 Maverick meta-llama/llama-4-maverick · tier 2 | 524K | $0.1875 | $0.6525 | Tools Vision | Unknown |
| OpenAI | GPT-5.4 nano openai/gpt-5.4-nano · tier 1 | 400K | $0.20 | $1.25 | Tools Vision | Unknown |
Gemini 3.5 Flash-Lite google/gemini-3.5-flash-lite · tier 1 | 1.05M | $0.30 | $2.50 | Tools Vision | Unknown | |
| Qwen | Qwen3.5 397B qwen/qwen3.5-397b-a17b · tier 3 | 262K | $0.50 | $3.60 | Tools Vision | Unknown |
| Mistral | Mistral Large 4 mistralai/mistral-large-4-0 · tier 3 | 1.05M | $0.70 | $2.10 | Tools Vision | Unknown |
| OpenAI | GPT-5.4 mini openai/gpt-5.4-mini · tier 2 | 400K | $0.75 | $4.50 | Tools Vision | Unknown |
Gemini 3.8 Flash google/gemini-3.8-flash · tier 3 | 1.05M | $0.825 | $4.125 | Tools Vision | Unknown | |
| Mistral | Mistral Medium 3.5 mistralai/mistral-medium-3-5 · tier 2 | 262K | $1.50 | $7.50 | Tools Vision | Unknown |
| DeepSeek | DeepSeek V4 Pro deepseek/deepseek-v4-pro · tier 3 | 1.05M | $1.60 | $3.20 | Tools | Unknown |
| OpenAI | GPT-5.3 Codex openai/gpt-5.3-codex · tier 3 | 400K | $1.75 | $14.00 | Tools Vision | Unknown |
| Anthropic | Claude Sonnet 5.5 anthropic/claude-sonnet-5.5 · tier 3 | 1M | $2.00 | $10.00 | Tools Vision | Unknown |
| OpenAI | GPT-5.4 openai/gpt-5.4 · tier 3 | 1.05M | $2.50 | $15.00 | Tools Vision | Unknown |
| Anthropic | Claude Opus 5.5 anthropic/claude-opus-5.5 · tier 4 | 1M | $4.00 | $20.00 | Tools Vision | Unknown |
| OpenAI | GPT-5.5 openai/gpt-5.5 · tier 4 | 1.05M | $5.00 | $30.00 | Tools Vision | Unknown |
Prices from OpenRouter public listing, fetched Nov 4, 2026. USD per million tokens, before Orbit's platform fee. Availability is unknown while the API is offline — a model is only usable when the operator has configured a key for its provider.
Methodology
Where these numbers come from.
Prices are synced, not typed
pnpm models:sync fetches OpenRouter's public model listing and writes a snapshot with per-token prices, context windows, output limits and a timestamp. The snapshot this site was built with was fetched on Nov 4, 2026. When the API is reachable the table above shows the API's current catalog instead.
What you are charged
Provider cost is computed from the provider-reported token usage at these per-token prices, then Orbit's platform fee is added (the currently configured default is 5% on inference). Amounts are kept in integer nano-dollars and rounded up to the nearest nano-dollar. See pricing.
Availability
A model is only usable if the operator has configured an API key for its provider (served direct) or an OpenRouter key (served via OpenRouter). Models without either show as Not configured — Orbit never pretends otherwise. A deterministic sandbox model exists for local development only and can never be enabled in production.
Tiers are routing hints, not benchmarks
Each model is assigned a tier and strengths (general, coding, reasoning, long_context, vision, fast) in Orbit's configuration. They steer Orbit Auto; they are not a claim about benchmark results.
- Tier 1
- Fast, inexpensive models for simple or high-volume tasks.
- Tier 2
- General-purpose models with solid coding and writing ability.
- Tier 3
- Strong reasoning and coding models for complex work.
- Tier 4
- The most capable models in the catalog, priced accordingly.
curl https://<your-orbit-host>/v1/models