Print-ready live report
Kilo intelligence
Benchmark signal / live operational data

AI model
performance
report

A concise view of coding completion, cost efficiency, model popularity, and inference provider reliability across the Kilo ecosystem.

Published
November 15, 2026
Benchmark
Terminal Bench 2.0
Inference window
Latest weekly snapshot
KILO AI MODEL REPORTNovember 15, 2026 / 01
01 / Capability and cost

KiloBench leaders

02

Models are ranked by completion rate on Terminal Bench 2.0. Cost per attempt is the observed average for a complete benchmark attempt, including unsuccessful attempts.

Top completion
79.3%
GPT-6 Astra ($$$$)
Top model cost
$107.29
Average cost per attempt
Lowest top-10 cost
$2.58
Among ranked models shown
Average completion
74.5%
Mean across the top 10
#ModelProviderCompletion rateCost / attempt
1GPT-6 Astra ($$$$)
5 attempts
openai79.3%$107.29
2GPT-5.6 Sol
5 attempts
openai76.2%$87.41
3Claude Fable 5.1 ($$$$)
5 attempts
anthropic76.2%$91.41
4DeepSeek V4.1 Flash
5 attempts
deepseek75.3%$2.58
5Gemini 3.8 Flash
5 attempts
google75.3%$112.27
6GPT-5.5
5 attempts
openai74.2%$72.63
7Grok 4.6
5 attempts
x-ai73.0%$33.83
8Kimi K3
5 attempts
moonshotai72.8%$48.38
9Claude Opus 5
5 attempts
anthropic71.5%$113.54
10Claude Fable 5 ($$$$)
5 attempts
anthropic71.0%$87.52
How to read this page. Completion rate is the KiloBench score: the share of benchmark tasks completed successfully. Published results require at least five attempts and an observed attempt cost; inactive and stealth models are excluded. Rankings prioritize higher completion, then lower cost when scores tie.
KILO AI MODEL REPORTNovember 15, 2026 / 02
02 / Real-world demand

Inference market view

03

The 20 most-used models are ranked by token volume from the latest Kilo Gateway weekly snapshot. Primary-provider metrics describe the provider carrying the largest traffic share for each model.

Observed volume
7.1T
Tokens across published records
Models ranked
20
Ordered by usage volume
Providers observed
53
Distinct inference providers
Most-used model
#01
nemotron-3.5-super-vl:free
#ModelPrimary providerUsage volumeCost / requestSuccess rate
1nemotron-3.5-super-vl:free
1 provider
NVIDIA2.1T$095.85%
2nemotron-3-ultra-550b-a55b:free
1 provider
NVIDIA1.6T$094.81%
3ernie-5.5-flash-preview:free
1 provider
Baidu853.9B$096.24%
4dots-3-note-preview:free
1 provider
Atlas Cloud504.4B$087.22%
5deepseek-v4.1-flash-1017
11 providers
Novita AI490.1B$0.002897.26%
6glyph-cluster
1 provider
Stealth314.1B$095.33%
7beam-free
1 provider
Reflection201.6B$094.28%
8nemotron-3.5-lightning:free
1 provider
NVIDIA130.6B$085.62%
9particle-1:free
1 provider
Particle AI129.2B$097.23%
10openrouter/free
11 providers
NVIDIA108.4B$087.42%
Primary provider. The provider with the highest token volume for a model. Success rate equals 100% minus observed error rate. Costs are average observed request costs during the reporting period; zero indicates free traffic in the source data.
KILO AI MODEL REPORTNovember 15, 2026 / 03
02 / Real-world demand

Inference market view / continued

04

The 20 most-used models are ranked by token volume from the latest Kilo Gateway weekly snapshot. Primary-provider metrics describe the provider carrying the largest traffic share for each model.

#ModelPrimary providerUsage volumeCost / requestSuccess rate
11deepseek-v4.1-flash
4 providers
Novita AI76.9B$0.002599.77%
12claude-opus-5.5
4 providers
Google74.4B$0.1899.50%
13north-mini-code:free
1 provider
Cohere64.6B$081.63%
14gpt-6-luna
4 providers
OpenAI61.8B$0.003799.95%
15glm-5.3-flash
7 providers
FriendliAI56.4B$0.006299.05%
16gpt-6.1-terra
3 providers
OpenAI39.2B$0.019299.97%
17gpt-6.1-sol
4 providers
OpenAI33.4B$0.057599.95%
18claude-sonnet-5.5
4 providers
Google29.2B$0.066299.15%
19nemotron-3-nano-omni-30b-a3b-reasoning:free
1 provider
NVIDIA26.6B$078.26%
20crouton-alpha
1 provider
Stealth19.7B$096.62%
Primary provider. The provider with the highest token volume for a model. Success rate equals 100% minus observed error rate. Costs are average observed request costs during the reporting period; zero indicates free traffic in the source data.
KILO AI MODEL REPORTNovember 15, 2026 / 04
03 / Provider routing detail

Cost and reliability

05

Provider-level traffic share, average request cost, and success rate for the ten most-used models. Traffic share uses the percentage published by Kilo Gateway.

01

nemotron-3.5-super-vl:free

2.1T tokens
ProviderTraffic shareCost / requestSuccess rate
NVIDIA100.00%$095.85%
02

nemotron-3-ultra-550b-a55b:free

1.6T tokens
ProviderTraffic shareCost / requestSuccess rate
NVIDIA100.00%$094.81%
03

ernie-5.5-flash-preview:free

853.9B tokens
ProviderTraffic shareCost / requestSuccess rate
Baidu100.00%$096.24%
04

dots-3-note-preview:free

504.4B tokens
ProviderTraffic shareCost / requestSuccess rate
Atlas Cloud100.00%$087.22%
KILO AI MODEL REPORTNovember 15, 2026 / 05
03 / Provider routing detail

Cost and reliability

06

Provider-level traffic share, average request cost, and success rate for the ten most-used models. Traffic share uses the percentage published by Kilo Gateway.

05

deepseek-v4.1-flash-1017

490.1B tokens
ProviderTraffic shareCost / requestSuccess rate
Novita AI70.25%$0.002897.26%
Alibaba23.98%$0.006099.83%
Baseten3.44%$0.004299.97%
Together AI1.49%$0.0033100.00%
Fireworks AI0.42%$0.009199.84%
DeepInfra0.23%$0.0004199.99%
Parasail0.05%$0.009999.94%
Subconscious0.04%$0.0056100.00%
DigitalOcean0.04%$0.014699.92%
Relace0.04%$0.002199.79%
Wafer0.02%$0.0166100.00%
06

glyph-cluster

314.1B tokens
ProviderTraffic shareCost / requestSuccess rate
Stealth100.00%$095.33%
KILO AI MODEL REPORTNovember 15, 2026 / 06
03 / Provider routing detail

Cost and reliability

07

Provider-level traffic share, average request cost, and success rate for the ten most-used models. Traffic share uses the percentage published by Kilo Gateway.

07

beam-free

201.6B tokens
ProviderTraffic shareCost / requestSuccess rate
Reflection100.00%$094.28%
08

nemotron-3.5-lightning:free

130.6B tokens
ProviderTraffic shareCost / requestSuccess rate
NVIDIA100.00%$085.62%
09

particle-1:free

129.2B tokens
ProviderTraffic shareCost / requestSuccess rate
Particle AI100.00%$097.23%
KILO AI MODEL REPORTNovember 15, 2026 / 07
03 / Provider routing detail

Cost and reliability

08

Provider-level traffic share, average request cost, and success rate for the ten most-used models. Traffic share uses the percentage published by Kilo Gateway.

10

openrouter/free

108.4B tokens
ProviderTraffic shareCost / requestSuccess rate
NVIDIA53.05%$087.42%
Atlas Cloud15.41%$087.09%
Recursive10.40%$085.81%
Cohere6.30%$083.93%
Baidu5.70%$089.27%
Particle AI4.68%$091.59%
Liquid2.08%$060.62%
Phala1.27%$085.78%
Novita AI0.80%$085.58%
Thinking Machines0.16%$078.44%
Ant Digital0.15%$089.02%