CtrlAltDebrief

APEX Agents benchmark

33 AI models ranked on APEX Agents. Each row shows the model’s score next to its cheapest API price per 1M tokens where it has one, so you can weigh quality against cost.

Top model today

Gemini 3.5 Flash (Google) leads at 47%, from $1.50 per 1M input via Google.

Data via arXiv:2601.14242 →

Models ranked

33

on this benchmark

Top score

47%

current leader

Type

Single eval

benchmark

Updated

Oct 14, 2026

benchmark scores

Cost vs APEX Agents

31 priced models
Most attractive quadrant (cheap + high)
= reasoning model (extended thinking)
0%10%20%30%40%50%$0.03$0.1$0.3$1$3COST / 1M INPUT TOKENS (log) →APEX AGENTS SCORE ↑Gemini 3.5 FlashQwen3.8 Max (0902)Kimi K3GPT-5.6 TerraGPT-5.5GPT-5.6 LunaGLM 5.2GPT-5.4gpt-oss-120bgpt-oss-20b

Each labelled name links to that model’s page. Curated to the cost-efficiency frontier, top scorers and cheapest so labels stay readable; the full field is in the leaderboard table below (and via “select all”). Reasoning models (extended “thinking” before answering) carry a ring around the dot.

Models ranked on APEX Agents

highest score first · 33 models
Maker
#
1Gemini 3.5 FlashGoogle47%
$1.50
via Google
$9.00
via Google
2Qwen3.8 Max (0902)Alibaba42%
$2.00
Routed quote
$6.00
Routed quote
3Kimi K3Moonshot41%
$0.70
Routed quote
$12.50
via Sail
4GPT-5.6 TerraOpenAI39%
$2.00
via OpenAI
$12.00
via OpenAI
5Step 5 PreviewStepfun38%
n/a
n/a
6GPT-5.5OpenAI38%
$5.00
via OpenAI
$30.00
via OpenAI
7GPT-5.6 LunaOpenAI36%
$0.20
via OpenAI
$1.20
via OpenAI
8GLM 5.2Zhipu34%
$0.05
Routed quote
$1.98
via Scx AI
9GPT-5.4OpenAI33%
$2.50
via Databricks
$15.00
via OpenAI
10Claude Opus 4.6Anthropic33%reasoning: adaptive
$5.00
via Anthropic
$25.00
via Anthropic
11Gemini 3.1 Pro PreviewGoogle32%
$2.00
via Google
$12.00
via Google
12Apodex 1.1Apodex31%
n/a
n/a
13Kimi K2.6Moonshot28%
$0.65
via Wandb
$3.40
via Novita
14GPT-5.4 MiniOpenAI28%
$0.75
via Databricks
$4.50
via OpenAI
15Claude Sonnet 4.6Anthropic28%reasoning: adaptive
$3.00
via Databricks
$15.00
via Anthropic
16Gemini 3 FlashGoogle28%reasoning: true
$0.50
Routed quote
$3.00
Routed quote
17inclusionAI: Ling 3.0 Flash FinOther27%
$0.075
Routed quote
$0.22
Routed quote
18GPT-5.4 NanoOpenAI25%
$0.20
via Databricks
$1.25
via Databricks
19DeepSeek V4 Pro 0423DeepSeek24%
$0.435
via Tencent
$0.87
via Tencent
20Qwen3.7 PlusAlibaba22%
$0.282
via Aihubmix
$1.13
via Aihubmix
21Grok 4.3xAI17%
$1.25
via xAI
$2.50
via xAI
22Qwen3.5 397B A17BAlibaba15%
$0.164
via Aihubmix
$0.986
via Aihubmix
23Step 3.7 FlashOther15%
$0.20
via DeepInfra
$1.15
via DeepInfra
24DeepSeek V3.2DeepSeek15%reasoning: true
$0.259
Routed quote
$0.38
via DeepInfra
25GLM 5Zhipu14%
$0.60
via DeepInfra
$1.92
Routed quote
26Grok 4 20 0309xAI14%
$1.25
via xAI
$2.50
via xAI
27Gemini 3.1 Flash Lite PreviewGoogle12%
$0.25
via Google
$1.50
via Google
28Kimi K2.5Moonshot12%
$0.45
via DeepInfra
$2.25
via DeepInfra
29MiniMax M2.7MiniMax11%
$0.25
via DeepInfra
$0.55
via Pinstripes
30gpt-oss-120bOpenAI3%
$0.03
via Wandb
$0.17
via DeepInfra
31MiMo-V2.5-ProXiaomi2%
$0.435
Routed quote
$0.87
Routed quote
32Nvidia Nemotron 3 Super 120B A12bNVIDIA2%
$0.085
via DeepInfra
$0.40
via DeepInfra
33gpt-oss-20bOpenAI1%
$0.0145
via Darkbloom
$0.07
via Darkbloom
—Nemotron 3 120B A12bNVIDIA2%Artificial Analysis did not report this score in its latest run on Oct 14, 2026
$0.50
via Cloudflare
$1.50
via Cloudflare
On-demand listed rates, OpenRouter routed quotes and Vercel reseller quotes. Enterprise commitments can be materially lower. Cost per task uses best-effort token usage. An OpenRouter routed quote is the price of whichever upstream provider OpenRouter prefers today, so it moves day to day.

Frequently asked questions

What is the APEX Agents benchmark?

APEX (AI Productivity Index for Agents) is a benchmark of 480 long-horizon, cross-application tasks written by investment-banking analysts, management consultants and corporate lawyers, testing whether agents can autonomously complete realistic professional workflows using files and tools. Graded against task-specific rubrics.

Which AI model scores highest on APEX Agents?

Gemini 3.5 Flash (Google) leads with 47%, from $1.50 per 1M input tokens from Google.

How many models are ranked on APEX Agents?

33 models carry an APEX Agents score. Scores come from Artificial Analysis; where a model has a listed API price it is drawn from LiteLLM and OpenRouter, and models with no current price are listed without one.

Other benchmarks

compare the same models on a different eval
Intelligence Indexcomposite indexCoding Indexcomposite indexAgentic Indexcomposite indexGPQA DiamondleaderboardHumanity's Last ExamleaderboardMMLU-ProleaderboardMMMU-ProleaderboardLiveCodeBenchleaderboardSciCodeleaderboardTerminalBench HardleaderboardTerminalBench v2.1leaderboardτ²-benchleaderboardτ-bench BankingleaderboardIT-Bench SREleaderboardGDPvalleaderboardCritPtleaderboardOmniscienceleaderboardIFBenchleaderboardLong-Context Reasoningcomposite indexSWE-bench VerifiedleaderboardSWE-bench ProleaderboardAIME 2026leaderboardARC-AGI-2leaderboardAA-Briefcaseleaderboard
Benchmark scores via Artificial Analysis · scores updated Oct 14, 2026 · Listed prices from LiteLLM, OpenRouter and Vercel · catalogue refreshed Oct 14, 2026 · 24 prices older, back to Aug 21, 2026

Every weekday

AI moves fast. Here's your debrief.

News, analysis, tools, and more.

For people who build with AI