Supported Models

All open-weight models plus your own custom/self-trained GGUF models. Run with alphachat run <model>.

12 models

DeepSeek V4 Flashdeepseek-v4-flash

high demandMoE

Params

284B (13B active)

Context

128K

Run

alphachat run deepseek-v4-flash

Qwen 3.6 35Bqwen-3.6-35b

high demandMoE

Params

35B (3B active)

Context

128K

Run

alphachat run qwen-3.6-35b

Mixtral 8x7Bmixtral-8x7b

medium demandMoE

Params

46.7B (12.9B active)

Context

32K

Run

alphachat run mixtral-8x7b

Qwen3.5 122Bqwen3.5-122b

medium demandMoE

Params

122B (8B active)

Context

128K

Run

alphachat run qwen3.5-122b

Qwen 3.5 27Bqwen3.5-27b

low demandDense

Params

27B

Context

128K

Run

alphachat run qwen3.5-27b

Qwen 3.5 9Bqwen3.5-9b

low demandDense

Params

Context

32K

Run

alphachat run qwen3.5-9b

Qwen 3.5 4Bqwen3.5-4b

low demandDense

Params

Context

32K

Run

alphachat run qwen3.5-4b

Gemma4 E4Bgemma4-e4b

low demandDense

Params

Context

32K

Run

alphachat run gemma4-e4b

Qwen3 14Bqwen3-14b

low demandDense

Params

14B

Context

32K

Run

alphachat run qwen3-14b

DeepSeek-R1 14Bdeepseek-r1-14b

low demandDense

Params

14B

Context

128K

Run

alphachat run deepseek-r1-14b

GPT-OSS 20Bgpt-oss-20b

low demandMoE

Params

20B (5B active)

Context

32K

Run

alphachat run gpt-oss-20b

Qwen1.5 MoE A2.7Bqwen1.5-moe

low demandMoE

Params

14B (2.7B active)

Context

32K

Run

alphachat run qwen1.5-moe

AlphaLlama downloads and optimizes models automatically.