๐Ÿ† Free Model Leaderboard

The best $0 OpenRouter models, ranked monthly on a fixed battery โ€” coding, reasoning, concision.

Latest run: August 2026 (2026-08-10) ยท source: OpenRouter :free catalog ยท auto-refreshed 1st of each month
#ModelCompositeA ยท CodingB ยท ReasoningC ยท Concisionฮ”Tokenstok/s
1Nemotron 3 Ultra (550B-A55B)9.99.810.010.0๐Ÿ†•208617.1
2Gemma 4 26B (A4B)9.810.09.410.0๐Ÿ†•139223.4
3Laguna S 2.19.19.89.48.0๐Ÿ†•90828.2
4Laguna XS 2.1 DNF A4.00.010.010.0๐Ÿ†•58491.5
5Ling 3.0 Tiny DNF B0.40.00.08.0๐Ÿ†•1825153.7
6Gemma 4 31B DNF A,B,C0.00.00.00.0๐Ÿ†•00

Composite = 0.40ร—A + 0.30ร—B + 0.30ร—C ยท ฮ” vs previous month (โ–ฒ/โ–ผ on ยฑ0.3) ยท โš ๏ธ DNF = task failed, โˆ’2.0 penalty ยท all runs at temperature 0, $0 free tier

๐Ÿ“ˆ Month-over-month history

MonthNemotron 3 Ultra (550B-A55B)Gemma 4 26B (A4B)Laguna S 2.1Laguna XS 2.1Ling 3.0 TinyGemma 4 31B
August 20269.9 ๐Ÿ†•9.8 ๐Ÿ†•9.1 ๐Ÿ†•4.0 ๐Ÿ†•0.4 ๐Ÿ†•0.0 ๐Ÿ†•

๐ŸŒ Model context & external signals

Vendor descriptions & release dates from the OpenRouter catalog; community traction from HuggingFace. Our composite is independent โ€” this is the outside view.

#1 NVIDIA: Nemotron 3 Ultra (free)9.9
1000K context ยท released Jun 2026 ยท HuggingFace 494K downloads ยท 312 likes

NVIDIA Nemotron 3 Ultra is an open frontier-reasoning and orchestration model from NVIDIA, with 55B active parameters out of 550B total (MoE). Built on a hybrid Transformer-Mamba mixture-of-experts architecture, it...

#2 Google: Gemma 4 26B A4B (free)9.8
262K context ยท released Apr 2026 ยท HuggingFace 10.2M downloads ยท 1366 likes

Gemma 4 26B A4B IT is an instruction-tuned Mixture-of-Experts (MoE) model from Google DeepMind. Despite 25.2B total parameters, only 3.8B activate per token during inference โ€” delivering near-31B quality at...

#3 Poolside: Laguna S 2.1 (free)9.1
262K context ยท released Jul 2026

Laguna S 2.1 is the latest coding agent model from Poolside. Laguna S 2.1 is a 118B total parameter model with 8B active parameters, scoring 70.2% on Terminal-Bench 2.1 and...

#4 Poolside: Laguna XS 2.1 (free)4.0
262K context ยท released Jul 2026

Laguna XS 2.1 is the latest coding agent model in the 33B-A3B category from Poolside and a step forward from their Laguna XS.2 model (released in April 2026). It combines...

#5 inclusionAI: Ling 3.0 Tiny (free)0.4
262K context ยท released Aug 2026

Ling 3.0 Tiny is a mixture-of-experts model from InclusionAI, with 1.3B active parameters out of 7.9B total. It is designed for responsive agents, instruction following, and multi-turn conversations, with switchable...

#6 Google: Gemma 4 31B (free)0.0
262K context ยท released Apr 2026 ยท HuggingFace 10.1M downloads ยท 3493 likes

Gemma 4 31B Instruct is Google DeepMind's 30.7B dense multimodal model supporting text and image input with text output. Features a 256K token context window, configurable thinking/reasoning mode, native function...

๐Ÿงช How it's scored

โ† back to imagine.josephwwu.com