What is the best llm for agents right now?

Gemini 3 Flash is the current #1 on this list. Rankings move when daily ingest updates SWE-bench, Elo, price, or latency.

How is this list ranked?

Coding + preference Elo for computer-use and multi-step tools. Weights: SWE-bench 35%, Elo 25%, Speed 15%, TTFT 10%, Out $ 15%.

Is preference Elo the only score?

Preference Elo is a crowd vote from LMArena / Arena. People see two hidden answers and pick the one they like more. The model that wins more often gets a higher Elo. That means people preferred it — not that it passed a school test. It is not SWE-bench, not accuracy, and not a number we invent. This list uses a named Stack Engine mix — Elo is only one input when the job is “best chatbot,” not when the job is “cheapest coding agent.”

Why should I trust this over a tweet?

Every cell has a source and an as-of date. We noindex thin pages and we do not fill missing scores. Open the model or pair page and read the snapshot.

Search Intent · best llm for agents

Best LLM for agents in 2026

SWE-bench and preference Elo for tool-using and computer-use stacks.

Weights:SWE-bench 35%Elo 25%Speed 15%TTFT 10%Out $ 15%

Current #1 Ranked PickScore 82.8 / 100

Gemini 3 Flash

Google · Closed flagship

SWE-bench: 75.8% (Top 13%)Elo: 1,495 (Top 34%)Speed: 190 tok/s (Top 4%)TTFT: 95 ms (Top 6%)

SWE-bench: 75.8%Elo: 1,495

View full model fact sheet

Complete Ranked Category List

Models ranked by verified benchmark weights across SWE-bench and coding preference evaluations.

#1Gemini 3 Flash(Google)

Score 82.8

SWE-bench: 75.8% (Top13%)Elo: 1,495 (Top34%)Speed: 190 tok/s (Top4%)TTFT: 95 ms (Top6%)

🏆 #1 Overall Leader in Category

#2Gemini 3.7 Flash(Google)

Score 78.9

SWE-bench: 72.4% (Top21%)Elo: 1,530 (Top23%)Speed: 215 tok/s (Top1%)TTFT: 82 ms (Top1%)

Rank #2vs #1

#3DeepSeek V4 Flash(DeepSeek)

Score 75.0

SWE-bench: 71.2% (Top26%)Elo: 1,490 (Top39%)Speed: 142 tok/s (Top19%)TTFT: 165 ms (Top25%)

Rank #3vs #1

#4Gemini 3.6 Flash(Google)

Score 73.4

SWE-bench: 70.8% (Top29%)Elo: 1,506 (Top32%)Speed: 198 tok/s (Top2%)TTFT: 92 ms (Top4%)

Rank #4vs #1

#5GLM-5.3(Zhipu)

Score 72.3

SWE-bench: 76.4% (Top9%)Elo: 1,558 (Top14%)Speed: 90 tok/s (Top56%)TTFT: 255 ms (Top63%)

Rank #5vs #1

#6Gemini 3.6 Pro(Google)

Score 72.3

SWE-bench: 74.6% (Top15%)Elo: 1,570 (Top9%)Speed: 102 tok/s (Top43%)TTFT: 205 ms (Top43%)

Rank #6vs #1

#7OpenAI o3-mini(OpenAI)

Score 72.0

SWE-bench: 78.5% (Top4%)Elo: 1,560 (Top12%)Speed: 92 tok/s (Top51%)TTFT: 290 ms (Top75%)

Rank #7vs #1

#8Claude Sonnet 5(Anthropic)

Score 71.9

SWE-bench: 73.8% (Top17%)Elo: 1,556 (Top16%)Speed: 118 tok/s (Top30%)TTFT: 175 ms (Top30%)

Rank #8vs #1

#9GPT-5.6 Terra(OpenAI)

Score 67.5

SWE-bench: 70.4% (Top30%)Elo: 1,548 (Top19%)Speed: 128 tok/s (Top26%)TTFT: 160 ms (Top23%)

Rank #9vs #1

#10Grok 4.6(xAI)

Score 67.2

SWE-bench: 69.1% (Top36%)Elo: 1,592 (Top6%)Speed: 118 tok/s (Top30%)TTFT: 195 ms (Top40%)

Rank #10vs #1

#11GPT-5.6 Sol(OpenAI)

Score 67.2

SWE-bench: 77.6% (Top6%)Elo: 1,608 (Top4%)Speed: 84 tok/s (Top66%)TTFT: 270 ms (Top65%)

Rank #11vs #1

#12Claude Opus 5(Anthropic)

Score 66.9

SWE-bench: 79.2% (Top2%)Elo: 1,624 (Top1%)Speed: 72 tok/s (Top76%)TTFT: 310 ms (Top78%)

Rank #12vs #1

Frequently asked questions

Plain-English methodology and leaderboard answers

: Gemini 3 Flash is the current #1 on this list. Rankings move when daily ingest updates SWE-bench, Elo, price, or latency.

What is preference Elo? · How rankings update

Search Intent · best llm for agents

Best LLM for agents in 2026

SWE-bench and preference Elo for tool-using and computer-use stacks.

Weights:SWE-bench 35%Elo 25%Speed 15%TTFT 10%Out $ 15%

Current #1 Ranked PickScore 82.8 / 100

Gemini 3 Flash

Google · Closed flagship

SWE-bench: 75.8% (Top 13%)Elo: 1,495 (Top 34%)Speed: 190 tok/s (Top 4%)TTFT: 95 ms (Top 6%)

SWE-bench: 75.8%Elo: 1,495

View full model fact sheet

Complete Ranked Category List

Models ranked by verified benchmark weights across SWE-bench and coding preference evaluations.

#1Gemini 3 Flash(Google)

Score 82.8

SWE-bench: 75.8% (Top13%)Elo: 1,495 (Top34%)Speed: 190 tok/s (Top4%)TTFT: 95 ms (Top6%)

🏆 #1 Overall Leader in Category

#2Gemini 3.7 Flash(Google)

Score 78.9

SWE-bench: 72.4% (Top21%)Elo: 1,530 (Top23%)Speed: 215 tok/s (Top1%)TTFT: 82 ms (Top1%)

Rank #2vs #1

#3DeepSeek V4 Flash(DeepSeek)

Score 75.0

SWE-bench: 71.2% (Top26%)Elo: 1,490 (Top39%)Speed: 142 tok/s (Top19%)TTFT: 165 ms (Top25%)

Rank #3vs #1

#4Gemini 3.6 Flash(Google)

Score 73.4

SWE-bench: 70.8% (Top29%)Elo: 1,506 (Top32%)Speed: 198 tok/s (Top2%)TTFT: 92 ms (Top4%)

Rank #4vs #1

#5GLM-5.3(Zhipu)

Score 72.3

SWE-bench: 76.4% (Top9%)Elo: 1,558 (Top14%)Speed: 90 tok/s (Top56%)TTFT: 255 ms (Top63%)

Rank #5vs #1

#6Gemini 3.6 Pro(Google)

Score 72.3

SWE-bench: 74.6% (Top15%)Elo: 1,570 (Top9%)Speed: 102 tok/s (Top43%)TTFT: 205 ms (Top43%)

Rank #6vs #1

#7OpenAI o3-mini(OpenAI)

Score 72.0

SWE-bench: 78.5% (Top4%)Elo: 1,560 (Top12%)Speed: 92 tok/s (Top51%)TTFT: 290 ms (Top75%)

Rank #7vs #1

#8Claude Sonnet 5(Anthropic)

Score 71.9

SWE-bench: 73.8% (Top17%)Elo: 1,556 (Top16%)Speed: 118 tok/s (Top30%)TTFT: 175 ms (Top30%)

Rank #8vs #1

#9GPT-5.6 Terra(OpenAI)

Score 67.5

SWE-bench: 70.4% (Top30%)Elo: 1,548 (Top19%)Speed: 128 tok/s (Top26%)TTFT: 160 ms (Top23%)

Rank #9vs #1

#10Grok 4.6(xAI)

Score 67.2

SWE-bench: 69.1% (Top36%)Elo: 1,592 (Top6%)Speed: 118 tok/s (Top30%)TTFT: 195 ms (Top40%)

Rank #10vs #1

#11GPT-5.6 Sol(OpenAI)

Score 67.2

SWE-bench: 77.6% (Top6%)Elo: 1,608 (Top4%)Speed: 84 tok/s (Top66%)TTFT: 270 ms (Top65%)

Rank #11vs #1

#12Claude Opus 5(Anthropic)

Score 66.9

SWE-bench: 79.2% (Top2%)Elo: 1,624 (Top1%)Speed: 72 tok/s (Top76%)TTFT: 310 ms (Top78%)

Rank #12vs #1

Frequently asked questions

Plain-English methodology and leaderboard answers

: Gemini 3 Flash is the current #1 on this list. Rankings move when daily ingest updates SWE-bench, Elo, price, or latency.

What is preference Elo? · How rankings update