SShogo Bench

DeepSeek V4.1 Flash 1017 vs Kimi K3

DeepSeek V4.1 Flash 1017 leads Kimi K3 by 2.7 points on the Shogo Agent Index, but their ranges overlap, so treat it as close. DeepSeek V4.1 Flash 1017 wins 8 and Kimi K3 wins 3 of 12 shared benchmarks. DeepSeek V4.1 Flash 1017 is cheaper for agent workloads.

Add more models to this comparison →

Pillars

PillarDeepSeek V4.1 Flash 1017Kimi K3
Agentic coding6363
Long-horizon autonomy6665
Knowledge-work agents4942
Computer use4543
Security & context learning5453

Shared benchmarks

BenchmarkDeepSeek V4.1 Flash 1017Kimi K3
APEX-Agentsindex69.8%47.8%
ARC-AGI94.5%94.5%
ARC-AGI-278.1%60.4%
Chess Puzzles46.0%37.0%
CritPt24.0%23.4%
FrontierMath Tier 446.3%39.0%
FrontierMath Tiers 1-371.2%72.2%
GPQA Diamond93.2%93.1%
OTIS Mock AIME99.4%97.2%
SciCode55.9%59.5%
SimpleQA Verified42.5%50.6%
WeirdML v39.1%7.5%

Specs and price

DeepSeek V4.1 Flash 1017Kimi K3
Agent Index57.6 (#17)54.9 (#26)
Agent $/M$0.050$1.39
Input / output $/M$0.15 / $0.60$3.00 / $15.00
Context1M1M
ReleasedOct 17, 2026Jul 16, 2026
WeightsOpenOpen
API hosts5129

Don’t pick one. Route.

Shogo can run DeepSeek V4.1 Flash 1017 and Kimi K3 side by side and send each step of an agent to the one that’s cheaper for it.

Try both in Shogo