All Models

All models are free, verified working, and accessible through the same endpoint. Use GET /v1/models to get the latest list programmatically.

curl https://kymaapi.com/v1/models

Recommended Models ⭐

These models offer the best balance of quality, speed, and capabilities.

Best all-rounder. General tasks, coding, reasoning. 128K context. Ultra-fast.

model="llama-3.3-70b"

Best for coding. Code generation, math, multilingual. 32K context.

model="qwen-3-32b"

Best multimodal. Vision capable, Google’s newest. 128K context.

model="gemma-4-31b"

Highest quality. 235B params, complex reasoning. Ultra-fast.

model="qwen-3-235b-cerebras"

Model ID	Name	Context	Speed	Best For
`llama-3.3-70b`	Llama 3.3 70B	128K	⚡ Fast	General, code, reasoning
`llama-4-scout`	Llama 4 Scout 17B	512K	⚡ Fast	Long documents, analysis
`llama-3.1-8b`	Llama 3.1 8B	8K	⚡⚡ Fastest	Quick tasks, classification
`qwen-3-32b`	Qwen 3 32B	32K	⚡ Fast	Code, math, multilingual
`kimi-k2`	Kimi K2	128K	⚡ Fast	Agentic coding, tool use
`gpt-oss-120b`	GPT-OSS 120B	128K	Medium	General intelligence, writing
`gpt-oss-20b`	GPT-OSS 20B	128K	⚡ Fast	Fast general tasks
`gemma-4-31b`	Gemma 4 31B	128K	Medium	Multimodal, vision
`gemma-4-26b-moe`	Gemma 4 26B MoE	128K	⚡ Fast	Efficient inference
`gemini-3-flash`	Gemini 3 Flash	1M	Medium	Ultra-long context
`gemini-2.5-flash`	Gemini 2.5 Flash	1M	Medium	Long context
`gemma-3-27b`	Gemma 3 27B	128K	Medium	General
`llama-3.1-8b-cerebras`	Llama 3.1 8B	8K	⚡⚡ Fastest	Ultra-fast inference
`qwen-3-235b-cerebras`	Qwen 3 235B	32K	⚡ Fast	Complex reasoning

Models are updated regularly. New models are added as they become available from providers.