Compare / Google vs Meta
Google vs Meta: which AI lab is ahead?
Each lab's best model on sale goes head to head on every kind of work we measure, from our own benchmarks and licensed sources. 31 models in all; a lab is only as good as the model you would actually pick from it.
Kinds of work led by each lab's best model there, of the 10 both are measured on.
Results as of 19 November 2026 (catalogue release 2026-11-19-3608133e8331). Models on sale means listed on OpenRouter today.
Google or Meta? The short answer
Google is ahead: its best models beat Meta's in 6 of the 10 kinds of work both are measured on; Meta leads in 4.
- Pick Google for product listings, agents and tool use, reasoning and knowledge, writing, sticking to the facts and speed under pressure.
- Pick Meta for decks from an analysis, user surveys, marketing planning and professional work.
- On our overall leaderboard, Gemini 4 Argon is #8 (70 of 100) and Muse Wonder is #10 (66).
- Prices start at $0.042 per million tokens for Google (Gemma 3 4B) and $0.058 for Meta (Llama 3.1 8B Instruct).
Head to head
Each line names the model behind the number. Prices are blended, three input tokens to one output, at the developer's list price where we have read it; see every model's price. Intelligence Index from Artificial Analysis; overall score from our leaderboard.
Which lab is better for what
For each kind of work, each lab's best model there and its rank among every model on sale. Further right is better; the better rank is in green.
Every model, by intelligence and price
Each dot is a model on sale with an Intelligence Index from Artificial Analysis. Up is smarter; left is cheaper. Hover or tap a dot for its name.
GoogleMeta
On our overall leaderboard
The 10 of their models we rank overall: one score out of 100 across everything we publish (how the score works). Only models with enough of our own results are ranked.
- #8Gemini 4 Argon70
- #10Muse Wonder66
- #13Muse Ember61
- #14Gemini 4 Neon60
- #14Gemini 4 Hydrogen60
- #17Muse Spark 1.358
- #18Gemini 3.8 Flash57
- #20Gemini 4 Carbon50
- #23Gemini 3.1 Pro Preview44
- #35Gemini 3.5 Flash-Lite24
Benchmark by benchmark
The headline result of every benchmark both labs have models on: each lab's best model and its rank among the models on sale. Green marks the better rank.
| Measure | Meta | |
|---|---|---|
| Agents and tool use | ||
| Berkeley Function Calling Leaderboard (BFCL) V4 · Overall accuracy · BFCL% correct, higher is better | 56.2%Gemini 2.5 Flash · 8 of 36 | 31.9%Llama 3.3 70B Instruct · 24 of 36 |
| Artificial Analysis · Terminal-Bench 4.0% of tasks, higher is better | 57.1%Gemini 4 Argon · 5 of 107 | 52.0%Muse Wonder · 9 of 107 |
| Artificial Analysis · Terminal-Bench 2.1% of tasks, higher is better | 87.6%Gemini 3.8 Flash · 8 of 107 | 85.4%Muse Spark 1.3 · 10 of 107 |
| Artificial Analysis · τ-bench banking% of tasks, higher is better | 45.8%Gemini 3.8 Flash · 10 of 104 | 50.5%Muse Spark 1.3 · 2 of 104 |
| Vending-Bench 2 · Money after a yearUS dollars, higher is better | $5,396Gemini 3.5 Flash · 22 of 53 | $6,520Muse Spark 1.1 · 13 of 53 |
| Decks from an analysis | ||
| DeckBench · Deck ratingrating, higher is better | 1,079Gemini 4 Hydrogen · 21 of 35 | 1,204Muse Wonder · 15 of 35 |
| Marketing planning | ||
| ROASBench · Overall scorescore out of 100, higher is better | 51.4Gemini 3.8 Flash · 13 of 35 | 51.8Muse Ember · 10 of 35 |
| Product listings | ||
| CatalogBench · Reliably publish-ready% of products, higher is better | 78.6%Gemini 4 Neon · 1 of 35 | 66.1%Muse Wonder · 7 of 35 |
| CatalogBench · Reliably publish-ready · sales brief% of products, higher is better | 32.1%Gemini 4 Neon · 13 of 35 | 51.8%Muse Wonder · 7 of 35 |
| Professional work | ||
| APEX-Agents · Tasks passed% of tasks, higher is better | 67.8%Gemini 3.7 Flash · 4 of 37 | 57.8%Muse Spark 1.3 · 12 of 37 |
| GDP.pdf · Professional document tasks% of rubric, higher is better | 23.8%Gemini 3.7 Flash · 11 of 31 | 27.6%Muse Spark 1.3 · 5 of 31 |
| Reasoning and knowledge | ||
| Artificial Analysis · Artificial Analysis Intelligence Indexindex score, higher is better | 52.6Gemini 4 Argon · 5 of 185 | 52.4Muse Wonder · 6 of 185 |
| Artificial Analysis · Humanity's Last Exam% of questions, higher is better | 57.1%Gemini 4 Argon · 3 of 186 | 50.1%Muse Wonder · 10 of 186 |
| Artificial Analysis · GPQA Diamond% of questions, higher is better | 95.3%Gemini 3.8 Flash · 2 of 168 | 94.1%Muse Spark 1.3 · 5 of 168 |
| Speed under pressure | ||
| BulletBench · Ladder Elo · Bullet 60sladder Elo, higher is better | 1,068Gemini 3.8 Flash · 1 of 36 | 409Muse Wonder · 18 of 36 |
| Sticking to the facts | ||
| Vectara Hallucination Leaderboard · Hallucination rate · Vectara% of summaries, lower is better | 3.3%Gemini 2.5 Flash-Lite · 2 of 71 | 7.7%Llama 4 Scout · 26 of 71 |
| Arena (formerly LMArena) · Overall · Text factualityArena rating, higher is better | 1,511Gemini 4 Argon · 1 of 109 | 1,483Muse Spark 1.2 · 10 of 109 |
| SimpleQA Verified (Epoch AI) · Correct answers% of questions, higher is better | 73.5%Gemini 3.1 Pro Preview · 3 of 68 | 60.3%Muse Spark 1.2 · 15 of 68 |
| User surveys | ||
| SurveyBench · Analysis ratingrating, higher is better | 1,096Gemini 4 Hydrogen · 13 of 26 | 1,439Muse Wonder · 3 of 26 |
| Writing | ||
| Arena (formerly LMArena) · Overall · TextArena rating, higher is better | 1,525Gemini 4 Argon · 1 of 153 | 1,494Muse Spark 1.3 · 8 of 153 |
| UGI Leaderboard · Writing score · UGIscore out of 100, higher is better | 78.6Gemini 3.8 Flash · 1 of 124 | 26.7Llama 4 Scout · 110 of 124 |
Their models, one against one
Quick answers
Is Google or Meta better?
Google is ahead: its best models beat Meta's in 6 of the 10 kinds of work both are measured on; Meta leads in 4.
What is Google's best model?
Gemini 4 Argon: #8 on the Spring Prompt overall leaderboard and 52.6 on the Artificial Analysis Intelligence Index.
What is Meta's best model?
Muse Wonder: #10 on the Spring Prompt overall leaderboard and 52.4 on the Artificial Analysis Intelligence Index.
Which is better for agents and tool use, Google or Meta?
Google: its best model there, Gemini 4 Argon, ranks 5 of 159 models on sale, against 11 for Meta's Muse Wonder.
Which is better for reasoning and knowledge, Google or Meta?
Google: its best model there, Gemini 4 Argon, ranks 4 of 187 models on sale, against 8 for Meta's Muse Wonder.
Which is better for writing, Google or Meta?
Google: its best model there, Gemini 4 Argon, ranks 1 of 169 models on sale, against 7 for Meta's Muse Spark 1.3.
Which is cheaper, Google or Meta?
Google's cheapest model with results, Gemma 3 4B, costs $0.042 per million tokens (three input to one output), against $0.058 for Llama 3.1 8B Instruct from Meta.
Other labs compared
- Google vs OpenAIBest model against best model
- Anthropic vs GoogleBest model against best model
- Google vs xAIBest model against best model
- Meta vs OpenAIBest model against best model
- Anthropic vs MetaBest model against best model
- Meta vs xAIBest model against best model
- DeepSeek vs GoogleBest model against best model
- DeepSeek vs MetaBest model against best model
Which model is best for your product?
Benchmarks aren't your task. Run models from Google, Meta and the rest on your own prompt, and see the cheapest one that passes.