IFM Documentation
This site is a JavaScript app. Plain-text versions of every page are served as static files:
- Index: https://docs.ifm.ai/llms.txt
- All pages in one file: https://docs.ifm.ai/llms-full.txt
- One page:
https://docs.ifm.ai/docs/<slug>.md
For AI and CLI tools: a link of the form https://docs.ifm.ai/#/<slug> is the same page as https://docs.ifm.ai/docs/<slug>.md (Markdown) or https://docs.ifm.ai/<slug>/ (static HTML). The part after # is never sent to the server, so fetch the .md URL instead. Example: https://docs.ifm.ai/#/migration → https://docs.ifm.ai/docs/migration.md.
Get Started
- Overview — Model portfolio, featured models, and where to start.
https://docs.ifm.ai/docs/overview.md - Quickstart — Create an API key, set the base URL, and make your first chat completion.
https://docs.ifm.ai/docs/quickstart.md
Models
- Model overview — Every IFM model with context length, languages, and hosted-API availability.
https://docs.ifm.ai/docs/model-catalog.md - K2 Horizon — Frontier agentic reasoning series; six open-weight sizes, 375B served on the API.
https://docs.ifm.ai/docs/introduction.md - Jais 3 — Arabic-specialized agentic model on the K2 Horizon architecture; IFM/Jais-3-MoVA-36B-A4B on the API.
https://docs.ifm.ai/docs/jais-3.md - Jais ASR — Arabic speech recognition with dialect ID, code-switching, and diacritization.
https://docs.ifm.ai/docs/jais-asr.md - K2 V2 — Fully-open 70B research models (deprecated on the API; weights remain).
https://docs.ifm.ai/docs/k2-v2.md - Jais 2 — Arabic–English chat models, 70B and 8B (weights only).
https://docs.ifm.ai/docs/jais-2.md
Capabilities
- Reasoning — reasoning_effort levels, reasoning_content, and per-family support.
https://docs.ifm.ai/docs/reasoning.md - Function calling — OpenAI-schema tools, the tool loop, and tool_choice.
https://docs.ifm.ai/docs/custom-function-calling.md - Streaming — SSE deltas for content, reasoning, and tool calls.
https://docs.ifm.ai/docs/streaming.md - Structured output — JSON mode and JSON Schema via response_format.
https://docs.ifm.ai/docs/structured-output.md - Multi-turn conversations — Replaying assistant turns and reasoning_content verbatim.
https://docs.ifm.ai/docs/multi-turn.md - Session routing — X-Session-ID for KV-cache affinity.
https://docs.ifm.ai/docs/session-routing.md
API Reference
- Chat completions API — POST /v1/chat/completions — parameters and response schema.
https://docs.ifm.ai/docs/chat-completions.md - Transcriptions API — POST /v1/audio/transcriptions — multipart request, language codes, response, formats.
https://docs.ifm.ai/docs/transcription.md - Limits — Daily token cap per key and rate-limit headers.
https://docs.ifm.ai/docs/limits.md - Errors — Status codes, error types, and retry guidance.
https://docs.ifm.ai/docs/errors.md
Deploy
- Deploy K2 Horizon — Self-host with SGLang or vLLM.
https://docs.ifm.ai/docs/deployment.md - Deploy Jais 3 — Self-host with vLLM.
https://docs.ifm.ai/docs/deploy-jais-3.md
Harness
- OpenCode — Configure OpenCode with an IFM provider.
https://docs.ifm.ai/docs/opencode.md - OpenClaw — Register IFM as an OpenClaw model provider.
https://docs.ifm.ai/docs/openclaw.md - Pi — Configure the Pi coding agent for IFM.
https://docs.ifm.ai/docs/pi.md - Hermes — IFM provider plugin and config for Hermes Agent.
https://docs.ifm.ai/docs/hermes.md
Lifecycle
- Migration to K2 Horizon — Moving from K2 V2: model name, context, reasoning replay, tool formats.
https://docs.ifm.ai/docs/migration.md - Deprecation — Retired models and base URLs with replacements and dates.
https://docs.ifm.ai/docs/deprecation.md