Economy LLMs: under $0.02 a review call

Models: 24; 7 scored
Cheapest call: $0.0027; GPT-OSS 20B (Groq)
Dearest call: $0.0195; Qwen3.8 27B (Workers AI)
Best bugs caught: 67%; GPT-6 Luna
Newest: 2026-10-17; DeepSeek V4 Flash

Pricing

Economy-tier models: list prices per 1M tokens, one review call and reading a pull request.
				Cache write / 1M
GPT-OSS 20B (Groq)OpenAI · on Groq	$0.075	$0.30	–	–	$0.0027	$0.002
Gemma 4 26B A4B (Workers AI)Google · on Cloudflare Workers AI	$0.10	$0.30	–	–	$0.0033	$0.0026
Gemini 2.5 Flash-LiteGoogle	$0.10	$0.40	$0.01	–	$0.0036	$0.0026
Claude Haiku 5.5AnthropicAbove 100K in: $0.50 / $2.50	$0.10	$0.50	$0.01	$0.125	$0.0038	$0.0026
GPT-6 LunaOpenAIAbove 272K in: $0.20 / $0.75	$0.10	$0.50	$0.01	–	$0.0038	$0.0026
GLM-5.3 Flash (Workers AI)Z.ai · on Cloudflare Workers AI	$0.15	$0.50	–	–	$0.0051	$0.0039
GPT-OSS 120B (Groq)OpenAI · on Groq	$0.15	$0.60	–	–	$0.0054	$0.0039
GPT-5.6 LunaOpenAIAbove 272K in: $0.40 / $1.80	$0.20	$1.20	$0.02	–	$0.0081	$0.0052
GPT-5.4 NanoOpenAI	$0.20	$1.25	$0.02	–	$0.0083	$0.0052
Codestral 25.08Mistral	$0.30	$0.90	–	–	$0.01	$0.0079
Gemini 3.1 Flash-LiteGoogle	$0.25	$1.50	$0.025	–	$0.0102	$0.0066
DeepSeek FlashDeepSeekOff-peak $0.15 / $0.60 · peak 01:00–04:00 and 06:00–10:00 UTC weekdays	$0.30	$1.20	–	–	$0.0108	$0.0079
GPT-OSS 120B (Workers AI)OpenAI · on Cloudflare Workers AI	$0.35	$0.75	–	–	$0.011	$0.0092
MiMo-V2.6-Pro (OpenRouter)Xiaomi · on OpenRouter	$0.435	$0.87	–	–	$0.0135	$0.0114
Gemini 2.5 FlashGoogle	$0.30	$2.50	$0.03	–	$0.0139	$0.0079
Gemini 3.5 Flash-LiteGoogle	$0.30	$2.50	$0.03	–	$0.0139	$0.0079
Inkling SmallThinking Machines · on OpenRouter	$0.45	$1.20	–	–	$0.0147	$0.0118
DeepSeek V4 Flash (Workers AI)DeepSeek · on Cloudflare Workers AI	$0.44	$1.32	–	–	$0.0147	$0.0115
Nemotron 3 Super (Workers AI)NVIDIA · on Cloudflare Workers AI	$0.50	$1.50	–	–	$0.0167	$0.0131
Mistral Large 3Mistral · on OpenRouter	$0.50	$1.50	–	–	$0.0167	$0.0131
GLM-5.2Z.ai · on OpenRouter	$0.0341	$7.00	–	–	$0.0178	$0.0009
GLM-4.7Z.ai · on OpenRouter	$0.54	$1.98	–	–	$0.0189	$0.0142
Qwen3.8 27B (Workers AI)Qwen · on Cloudflare Workers AI	$0.45	$3.20	–	–	$0.0195	$0.0118
DeepSeek V4 FlashDeepSeek · retiredOff-peak $0.15 / $0.60 · peak 01:00–04:00 and 06:00–10:00 UTC weekdays	$0.30	$1.20	–	–	$0.0108	$0.0079

Benchmark

Benchmark results

1=	GPT-6 LunaOpenAI · Economy	67%	67%	100%	0.7	$0.0028	$0.0046	29s
1=	Claude Haiku 5.5Anthropic · Economy	64%	67%	98%	1.4	$0.0038	$0.0064	15s
2	GLM-5.3 Flash (Workers AI)Z.ai · Economy · open weights	61%	61%	100%	1.8	$0.0035	$0.0066	254s
5	DeepSeek V4 Flash (Workers AI)DeepSeek · Economy · open weights	49%	49%	81%	1.3	$0.0178	$0.0309	137s
8	GPT-OSS 120B (Workers AI)OpenAI · Economy · open weights	40%	40%	100%	2.5	$0.0075	$0.0189	55s
13	Gemma 4 26B A4B (Workers AI)Google · Economy · open weights	26%	27%	91%	0.4	$0.0048	$0.0209	308s
16	Gemini 3.5 Flash-LiteGoogle · Economy	11%	11%	98%	0.4	$0.0059	$0.0528	103s

Methodology

By language and bug kind

Share of known bugs each model caught, by language
Model	All	TypeScript6 bugs	Go4 bugs	Java4 bugs	Rust4 bugs	C#3 bugs	JavaScript3 bugs	Kotlin3 bugs	PHP3 bugs	Python3 bugs	Ruby3 bugs	Swift3 bugs	C++2 bugs	C1 bug
GPT-6 Luna	67%	83%	25%	65%	75%	56%	67%	67%	67%	100%	67%	100%	50%	0%
Claude Haiku 5.5	64%	83%	75%	40%	75%	67%	33%	67%	100%	39%	67%	67%	50%	0%
GLM-5.3 Flash (Workers AI)	61%	57%	31%	45%	88%	30%	67%	89%	33%	78%	67%	67%	100%	67%
DeepSeek V4 Flash (Workers AI)	49%	46%	25%	55%	46%	22%	56%	61%	78%	61%	44%	67%	50%	0%
GPT-OSS 120B (Workers AI)	40%	42%	13%	46%	71%	41%	11%	44%	33%	78%	33%	44%	33%	0%
Gemma 4 26B A4B (Workers AI)	26%	33%	0%	30%	38%	7%	0%	56%	33%	44%	67%	0%	0%	0%
Gemini 3.5 Flash-Lite	11%	33%	0%	5%	0%	0%	0%	33%	0%	17%	0%	0%	50%	0%

Quality against cost

Higher catches more · further left is cheaper

OpenAIAnthropicZ.aiDeepSeekGoogleFaded: models outside this list95% range: the line through each point

Every point as a table

Bugs caught and measured cost per review, by model
Model	Bugs caught	95% range	Cost / review
GPT-6 Luna	67%	53%–80%	$0.0028
Claude Haiku 5.5	64%	50%–78%	$0.0038
GLM-5.3 Flash (Workers AI)	61%	50%–72%	$0.0035
DeepSeek V4 Flash (Workers AI)	49%	37%–60%	$0.0178
GPT-OSS 120B (Workers AI)	40%	29%–52%	$0.0075
Gemma 4 26B A4B (Workers AI)	26%	15%–38%	$0.0048
Gemini 3.5 Flash-Lite	11%	3%–21%	$0.0059

Other benchmarks

Outside benchmarks by model
Model	Arena WebDev	Arena Agent	Epoch Capabilities Index
Claude Haiku 5.5	1585High · range 1573 to 1596	0.015High · range 0.005 to 0.024	159range 157 to 162
GPT-6 Luna	1578Max · range 1572 to 1584	0.030Max · range 0.025 to 0.035	156range 154 to 159
GPT-5.6 Luna	–	-0.021xHigh · range -0.028 to -0.013	156range 154 to 159
GPT-5.4 Nano	–	–	146range 144 to 149
Gemini 2.5 Flash	–	–	141range 139 to 142
Gemini 3.5 Flash-Lite	1441range 1401 to 1482	–	145range 143 to 147
Gemini 3.1 Flash-Lite	–	–	145range 142 to 146
MiMo-V2.6-Pro (OpenRouter)	1625range 1615 to 1635	0.040range 0.029 to 0.050	–
Mistral Large 3	1231range 1205 to 1256	–	–
Inkling Small	1409range 1400 to 1418	-0.151range -0.169 to -0.133	150range 147 to 152
GLM-5.2	1605Max · range 1599 to 1611	0.023Max · range 0.014 to 0.031	152range 150 to 154
GLM-4.7	1435range 1423 to 1447	–	144range 142 to 146

Source: LMArena CC BY 4.0, Oct 16, 2026 · Epoch AI CC BY 4.0, Oct 7, 2026

Related lists

Updated Oct 16, 2026All prices pricing.json