Search Intent · fastest llm
Fastest LLM in 2026 (TTFT and tok/s)
Time-to-first-token and output speed from dated snapshots. For voice, widgets, and tight loops.
Weights:Elo 35%TTFT 25%Speed 20%Out $ 20%
Current #1 Ranked PickScore 81.5 / 100
Gemini 3.7 Flash
Google · Closed flagship
Elo: 1,530 (Top 23%)TTFT: 82 ms (Top 1%)Speed: 215 tok/s (Top 1%)Out $: $1.875/1M (Top 50%)
SWE-bench: 72.4%Elo: 1,530
View full model fact sheetComplete Ranked Category List
Models ranked by verified benchmark weights across SWE-bench and coding preference evaluations.
Score 81.5
Elo: 1,530 (Top23%)TTFT: 82 ms (Top1%)Speed: 215 tok/s (Top1%)Out $: $1.875/1M (Top50%)
🏆 #1 Overall Leader in Category
Score 75.3
Elo: 1,466 (Top51%)TTFT: 88 ms (Top2%)Speed: 188 tok/s (Top5%)Out $: $0.6/1M (Top27%)
Rank #2vs #1
Score 74.8
Elo: 1,506 (Top31%)TTFT: 92 ms (Top4%)Speed: 198 tok/s (Top2%)Out $: $3.75/1M (Top65%)
Rank #3vs #1
Score 73.6
Elo: 1,495 (Top33%)TTFT: 95 ms (Top6%)Speed: 190 tok/s (Top4%)Out $: $3/1M (Top63%)
Rank #4vs #1
Score 73.3
Elo: 1,490 (Top38%)TTFT: 165 ms (Top25%)Speed: 142 tok/s (Top19%)Out $: $0.28/1M (Top17%)
Rank #5vs #1
Score 72.6
Elo: 1,475 (Top44%)TTFT: 150 ms (Top20%)Speed: 130 tok/s (Top24%)Out $: $0.14/1M (Top11%)
Rank #6vs #1
Score 67.4
Elo: 1,548 (Top19%)TTFT: 160 ms (Top23%)Speed: 128 tok/s (Top26%)Out $: $6/1M (Top75%)
Rank #7vs #1
Score 67.3
Elo: 1,360 (Top78%)TTFT: 110 ms (Top9%)Speed: 180 tok/s (Top8%)Out $: $0.1/1M (Top8%)
Rank #8vs #1
Score 66.9
Elo: 1,592 (Top6%)TTFT: 195 ms (Top40%)Speed: 118 tok/s (Top30%)Out $: $6/1M (Top75%)
Rank #9vs #1
Score 66.3
Elo: 1,472 (Top47%)TTFT: 120 ms (Top12%)Speed: 160 tok/s (Top11%)Out $: $2.5/1M (Top60%)
Rank #10vs #1
Score 66.0
Elo: 1,412 (Top69%)TTFT: 130 ms (Top14%)Speed: 155 tok/s (Top14%)Out $: $0.3/1M (Top18%)
Rank #11vs #1
Score 64.7
Elo: 1,468 (Top49%)TTFT: 170 ms (Top27%)Speed: 130 tok/s (Top24%)Out $: $0.8/1M (Top33%)
Rank #12vs #1
Frequently asked questions
Plain-English methodology and leaderboard answers
- Gemini 3.7 Flash is the current #1 on this list. Rankings move when daily ingest updates SWE-bench, Elo, price, or latency.