Spanish

15 of 15 models asked a question written entirely in Spanish, in its own script, answered it correctly. Answering in English is a failure here, not a pass — a model with no coverage cannot parse the question in the first place. Every verdict below is signed by the machine that took it.

Answered correctly

Best recall-proved context first — the order in which you would try them.

• hf.co/unsloth/gemma-4-E2B-it-GGUF:q4_k_m128K ctx · 291.2 tok/s · 13 machines
• hf.co/unsloth/gemma-4-E2B-it-GGUF:Q4_K_M128K ctx · 291.2 tok/s · 13 machines
• hf.co/unsloth/GLM-4.7-Flash-GGUF:Q4_K_M128K ctx · 213.3 tok/s · 14 machines
• hf.co/unsloth/gemma-4-E2B-it-GGUF:IQ4_XS128K ctx · 182.9 tok/s · 3 machines
• gemma4:e2b128K ctx · 170.5 tok/s · 10 machines
• hf.co/unsloth/gpt-oss-20b-GGUF:Q4_K_M128K ctx · 73.2 tok/s · 6 machines
• hf.co/unsloth/gemma-4-E2B-it-GGUF:Q8_0128K ctx · 67.0 tok/s · 2 machines
• hf.co/unsloth/gemma-4-E2B-it-GGUF:BF16128K ctx · 38.3 tok/s · 2 machines
• hf.co/unsloth/gemma-4-26B-A4B-it-GGUF:Q8_0128K ctx · 20.8 tok/s · 2 machines
• hf.co/unsloth/gemma-4-31B-it-GGUF:Q4_K_M128K ctx · 3.4 tok/s · 1 machine
• hf.co/unsloth/gpt-oss-20b-GGUF:Q3_K_S64K ctx · 54.7 tok/s · 3 machines
• hf.co/unsloth/Phi-4-mini-reasoning-GGUF:IQ3_XXS4K ctx · 189.4 tok/s · 6 machines
• hf.co/unsloth/gpt-oss-20b-GGUF:Q4_04K ctx · 4.6 tok/s · 3 machines
• hf.co/huihui-ai/Huihui-Qwen3.6-35B-A3B-abliterated-MTP-GGUF:Q6_K4K ctx · 4.3 tok/s · 2 machines
• hf.co/unsloth/Qwen3.5-35B-A3B-GGUF:Q4_K_M4K ctx · 4.2 tok/s · 2 machines

A model absent from both lists was never asked in Spanish, and that is not a claim either way. The probe set covers a fixed list of languages; a script missing from it is missing from the evidence, not from the model.