The AI model catalog, ranked by use case.

334 models across 57 vendors, ranked head-to-head for 82 real tasks. Updated daily from the OpenRouter catalog. Pricing, context, modality, tool support : all in one place.

334 models 82 task rankings 27 free models updated 2026-06-21
From the editors
2026-06-21

TheDrummer: Rocinante 12B - Solid Reasoning, Watch the Costs

TheDrummer: Rocinante 12B has been quietly outperforming its weight class on multi-step reasoning tasks. We've seen it handle complex SQL generation and data transformation pipelines where smaller models hallucinate table relationships. It's particularly good at maintaining context through 4-5 step instructions without needing excessive prompting. The 32k context window gives you room to dump schema definitions and example queries without token anxiety. Where it loses: creative writing and open-ended generation feel noticeably more formulaic than Llama 3 70B or Claude alternatives.

The pricing isn't published yet, which is the main gotcha. For production use, you need those numbers before committing to integration work. If TheDrummer prices this competitively with other 12B models, it's worth testing against your current solution for any structured reasoning tasks. If they're charging premium rates to match the quality uplift, you're probably better off jumping straight to a 70B model and getting broader capability coverage.

Test it on your gnarliest parsing task, but wait for pricing before refactoring anything.

Just released

auto-detected from the live catalog

Top-ranked available models

all 334 →

Browse by use case

all 82 tasks →

Cheapest paid models

see all →
#ModelIn / 1MOut / 1MContext
1 inclusionAI: Ling-2.6-flashinclusionai/ling-2.6-flash $0.010 $0.030 262,144
2 Meta: Llama 3.1 8B Instructmeta-llama/llama-3.1-8b-instruct $0.020 $0.030 131,072
3 Mistral: Mistral Nemomistralai/mistral-nemo $0.020 $0.030 131,072
4 Sao10K: Llama 3 8B Lunarissao10k/l3-lunaris-8b $0.040 $0.050 8,192
5 MythoMax 13Bgryphe/mythomax-l2-13b $0.060 $0.060 4,096
6 IBM: Granite 4.0 Microibm-granite/granite-4.0-h-micro $0.017 $0.112 131,000
7 Mistral: Mistral Small 3mistralai/mistral-small-24b-instruct-2501 $0.050 $0.080 32,768
8 Qwen: Qwen2.5 7B Instructqwen/qwen-2.5-7b-instruct $0.040 $0.100 131,072
9 LiquidAI: LFM2-24B-A2Bliquid/lfm-2-24b-a2b $0.030 $0.120 128,000
10 IBM: Granite 4.1 8Bibm-granite/granite-4.1-8b $0.050 $0.100 131,072