spark 1.37 . by forgewright-ai

~ > spark model list -- 26 models, one list

modelfileRAMlicensetestedgrounded
qwen3-1-7b1.0 GB3 GBApache 2.0yes
qwen3-4b2.3 GB5 GBApache 2.0yes5/9 2026-09-14
qwen3-8b4.7 GB7 GBApache 2.0yes21/27 2026-09-13
qwen3-14b8.4 GB11 GBApache 2.0yes
qwen3-30b-a3b17.4 GB21 GBApache 2.0yes
qwen3-4b-thinking2.3 GB5 GBApache 2.0
qwen3-coder-30b-a3b17.3 GB21 GBApache 2.0
qwen2-5-coder-7b4.4 GB7 GBApache 2.0
qwen2-5-7b4.4 GB7 GBApache 2.0
qwen2-5-14b8.4 GB11 GBApache 2.0
mistral-7b-v0-34.1 GB6 GBApache 2.0
mistral-nemo-12b7.0 GB10 GBApache 2.0
phi-4-mini2.3 GB5 GBMIT
phi-48.4 GB11 GBMIT
deepseek-r1-qwen-7b4.4 GB7 GBMIT
deepseek-r1-qwen-14b8.4 GB11 GBMIT
smollm2-1-7b1.0 GB3 GBApache 2.0
gpt-oss-20b10.8 GB14 GBApache 2.0
granite-4-2-8b5.0 GB7 GBApache 2.0yes25/30 2026-09-18
llama3-2-1b0.8 GB3 GBLlama 3.2 Community License
llama3-2-3b1.9 GB4 GBLlama 3.2 Community License
llama3-1-8b4.6 GB7 GBLlama 3.1 Community License
gemma3-1b0.8 GB3 GBGemma Terms of Use
gemma3-4b2.3 GB5 GBGemma Terms of Use
gemma3-12b6.8 GB9 GBGemma Terms of Use5/9 2026-09-14
gemma3-27b15.4 GB19 GBGemma Terms of Use

spark model NAME serves one; spark ember NAME adds a second, for conversations. auto picks among the rows tested on the line under Apache-2.0 or MIT; any other license is shown and asked about before the download. Every file is sha256-verified. Your own rows: spark model add URL --license. A row is a pull request: CONTRIBUTING.md.

source: models.env on GitHub -- this page is that file at spark v1.37