modelright
← Back to Task Picks

Coding Task Picks

Capable models suited for code generation, refactoring, and debugging — benchmarked models first.

Selection Rule: Available paid models carrying an Artificial Analysis coding index rank first, highest index wins; remaining slots fall back to the five most-listed active providers, sorted by output price ascending. Removed models excluded.
RankModelProviderContext WindowInput / 1MOutput / 1MAction
#1Mistral NemoMistral131,072 tokens$0.03$0.03View Details
#2Mistral Small 3Mistral32,768 tokens$0.05$0.08View Details
#3gpt-oss-20bOpenAI131,072 tokens$0.02$0.09View Details
#4Gemma 3 4BGoogle131,072 tokens$0.05$0.10View Details
#5Ministral 3 3B 2512Mistral131,072 tokens$0.10$0.10View Details
#6gpt-oss-20b (batch)OpenAI131,072 tokens$0.02$0.11View Details
#7Qwen3.7 FlashQwen1,000,000 tokens$0.03$0.13View Details
#8gpt-oss-120b (batch)OpenAI131,072 tokens$0.03$0.14View Details
#9Gemma 3 12BGoogle131,072 tokens$0.05$0.15View Details
#10Ministral 3 8B 2512Mistral262,144 tokens$0.15$0.15View Details