Skip to content

Listings

All listings: OpenRouter

Every listing for the models on page 15 of 26 of OpenRouter, including each model's other versions and routes. Back to OpenRouter

Sold by OpenRouter

48 listings on this page sold by OpenRouter.

Select a score for details.

About these columns
llminfo IQ
llminfo IQ (capability index). How llminfo IQ is computed
Coding IQ
llminfo Coding IQ (coding capability, its own scale). How Coding IQ is computed
Value
Value for money. 50 means as expected for its results. Above 50, cheaper per task. How Value for money works
Modelllminfo IQCoding IQInput $/1MOutput $/1MMax Input TokensValueEpoch Capabilities Index
Qwen2.5-VL 7B Instruct
🇨🇳
Maker headquartered in CN, according to the company's own website. The citation is on the model page.
100.7
From 78 benchmarks. Plenty of data behind it.
Not enough data yet$0.20$0.20Not published by the provider
39.1
Poor value (39.1). Costs more per task than its benchmark results predict. Little data behind it, so treat it as a rough estimate.
Not measured
Ministral 3B
🇫🇷
Maker headquartered in FR, according to the company's own website. The citation is on the model page.
Open weights. The weights are publicly available, according to the provider's model list or a public Hugging Face repo.
100.6
From 24 benchmarks. Plenty of data behind it.
Not enough data yet$0.04$0.04Not published by the provider
38.6
Poor value (38.6). Costs more per task than its benchmark results predict. Little data behind it, so treat it as a rough estimate.
Not measured
Ministral 3B
🇫🇷
Maker headquartered in FR, according to the company's own website. The citation is on the model page.
Open weights. The weights are publicly available, according to the provider's model list or a public Hugging Face repo.
mistralai/ministral-3b-2512 · 2512
100.6
From 24 benchmarks. Plenty of data behind it.
Not enough data yet$0.10cache read $0.01$0.10131K
38.6
Poor value (38.6). Costs more per task than its benchmark results predict. Little data behind it, so treat it as a rough estimate.
Not measured
Ministral 3B
🇫🇷
Maker headquartered in FR, according to the company's own website. The citation is on the model page.
Open weights. The weights are publicly available, according to the provider's model list or a public Hugging Face repo.
mistralai/ministral-3b-2512 · eu+2512
Inference by Mistral AI
100.6
From 24 benchmarks. Plenty of data behind it.
Not enough data yet$0.11cache read $0.011$0.11131K
38.6
Poor value (38.6). Costs more per task than its benchmark results predict. Little data behind it, so treat it as a rough estimate.
Not measured
Ministral 3B
🇫🇷
Maker headquartered in FR, according to the company's own website. The citation is on the model page.
Open weights. The weights are publicly available, according to the provider's model list or a public Hugging Face repo.
mistralai/ministral-3b-2512 · zdr+2512
Inference by Mistral AI
100.6
From 24 benchmarks. Plenty of data behind it.
Not enough data yet$0.10cache read $0.01$0.10131K
38.6
Poor value (38.6). Costs more per task than its benchmark results predict. Little data behind it, so treat it as a rough estimate.
Not measured
GPT-3.5 Turbo
🇺🇸
Maker headquartered in US, according to the company's own website. The citation is on the model page.
Newer version
Newer versionGPT-4 TurboWe compared the version names in our catalogue.
99.4
From 417 benchmarks. Plenty of data behind it.
≈ 76 ±11
Coding IQ 75.7, 95% interval 65–87. Little data behind it, so treat it as a rough estimate. Uses the model's best published results. It has its own scale, separate from llminfo IQ.
$0.50$1.50Not published by the provider
27.7
Poor value (27.7). Costs more per task than its benchmark results predict. A fair amount of data behind it.
112.9
Epoch Capabilities IndexSource: Epoch AI Benchmarking Hub API
GPT-3.5 Turbo
🇺🇸
Maker headquartered in US, according to the company's own website. The citation is on the model page.
Newer version
Newer versionGPT-4 TurboWe compared the version names in our catalogue.
gpt-3.5-turbo-1106 · 1106
99.4
From 417 benchmarks. Plenty of data behind it.
≈ 76 ±11
Coding IQ 75.7, 95% interval 65–87. Little data behind it, so treat it as a rough estimate. Uses the model's best published results. It has its own scale, separate from llminfo IQ.
$1.00$2.00Not published by the provider
27.7
Poor value (27.7). Costs more per task than its benchmark results predict. A fair amount of data behind it.
112.9
Epoch Capabilities IndexSource: Epoch AI Benchmarking Hub API
GPT-3.5 Turbo
🇺🇸
Maker headquartered in US, according to the company's own website. The citation is on the model page.
Newer version
Newer versionGPT-4 TurboWe compared the version names in our catalogue.
openai/gpt-3.5-turbo-0613 · 0613
99.4
From 417 benchmarks. Plenty of data behind it.
≈ 76 ±11
Coding IQ 75.7, 95% interval 65–87. Little data behind it, so treat it as a rough estimate. Uses the model's best published results. It has its own scale, separate from llminfo IQ.
$1.00$2.004K
27.7
Poor value (27.7). Costs more per task than its benchmark results predict. A fair amount of data behind it.
112.9
Epoch Capabilities IndexSource: Epoch AI Benchmarking Hub API
GPT-3.5 Turbo
🇺🇸
Maker headquartered in US, according to the company's own website. The citation is on the model page.
Newer version
Newer versionGPT-4 TurboWe compared the version names in our catalogue.
openai/gpt-3.5-turbo-0613 · 0613
Inference by Microsoft Azure
99.4
From 417 benchmarks. Plenty of data behind it.
≈ 76 ±11
Coding IQ 75.7, 95% interval 65–87. Little data behind it, so treat it as a rough estimate. Uses the model's best published results. It has its own scale, separate from llminfo IQ.
$1.00$2.004K
27.7
Poor value (27.7). Costs more per task than its benchmark results predict. A fair amount of data behind it.
112.9
Epoch Capabilities IndexSource: Epoch AI Benchmarking Hub API
DeepHermes 3 Mistral 24B Preview
🇺🇸
Maker headquartered in US, according to the company's own website. The citation is on the model page.
99.2
From 6 benchmarks. A fair amount of data behind it.
Not enough data yet$0.15$0.59Not published by the providerNot enough data yetNot measured
Llama 2 70B Chat
🇺🇸
Maker headquartered in US, according to the company's own website. The citation is on the model page.
98.8
From 50 benchmarks. A fair amount of data behind it.
Not enough data yet$0.90$0.90Not published by the providerNot enough data yetNot measured
Qwen2.5 7B Instruct
🇨🇳
Maker headquartered in CN, according to the company's own website. The citation is on the model page.
Open weights. The weights are publicly available, according to the provider's model list or a public Hugging Face repo.
98.8
From 248 benchmarks. Plenty of data behind it.
73 ±9stale
Coding IQ 73.2, 95% interval 64–82. A fair amount of data behind it. Uses the model's best published results. It has its own scale, separate from llminfo IQ. Its newest result is over a year old.
$0.04$0.10Not published by the provider
54.3
Good value (54.3). Cheaper per task than its benchmark results predict. Little data behind it, so treat it as a rough estimate.
Not measured
Jamba 1.5 Mini
🇮🇱
Maker headquartered in IL, according to the company's own website. The citation is on the model page.
98.6
From 113 benchmarks. Plenty of data behind it.
Not enough data yet$0.20$0.40Not published by the provider
26.9
Poor value (26.9). Costs more per task than its benchmark results predict. A fair amount of data behind it.
Not measured
Mistral Nemo
🇫🇷
Maker headquartered in FR, according to the company's own website. The citation is on the model page.
Open weights. The weights are publicly available, according to the provider's model list or a public Hugging Face repo.
mistralai/mistral-nemo:free · free
98.1
From 85 benchmarks. Plenty of data behind it.
Not enough data yet$0.00$0.00Not published by the provider
52.2
Fair value (52.2). Costs per task about what its benchmark results predict. Little data behind it, so treat it as a rough estimate.
113.9
Epoch Capabilities IndexSource: Epoch AI Benchmarking Hub API
Mistral Nemo
🇫🇷
Maker headquartered in FR, according to the company's own website. The citation is on the model page.
Open weights. The weights are publicly available, according to the provider's model list or a public Hugging Face repo.
98.1
From 85 benchmarks. Plenty of data behind it.
Not enough data yet$0.029$0.03131K
52.2
Fair value (52.2). Costs per task about what its benchmark results predict. Little data behind it, so treat it as a rough estimate.
113.9
Epoch Capabilities IndexSource: Epoch AI Benchmarking Hub API
Mistral Nemo
🇫🇷
Maker headquartered in FR, according to the company's own website. The citation is on the model page.
Open weights. The weights are publicly available, according to the provider's model list or a public Hugging Face repo.
fp8
Inference by DeepInfra
98.1
From 85 benchmarks. Plenty of data behind it.
Not enough data yet$0.029$0.03131K
52.2
Fair value (52.2). Costs per task about what its benchmark results predict. Little data behind it, so treat it as a rough estimate.
113.9
Epoch Capabilities IndexSource: Epoch AI Benchmarking Hub API
Mistral Nemo
🇫🇷
Maker headquartered in FR, according to the company's own website. The citation is on the model page.
Open weights. The weights are publicly available, according to the provider's model list or a public Hugging Face repo.
fp8/flex · fp8
Inference by DeepInfra
98.1
From 85 benchmarks. Plenty of data behind it.
Not enough data yet$0.0232$0.024131K
52.2
Fair value (52.2). Costs per task about what its benchmark results predict. Little data behind it, so treat it as a rough estimate.
113.9
Epoch Capabilities IndexSource: Epoch AI Benchmarking Hub API
Mistral Nemo
🇫🇷
Maker headquartered in FR, according to the company's own website. The citation is on the model page.
Open weights. The weights are publicly available, according to the provider's model list or a public Hugging Face repo.
fp8
Inference by Dekallm
98.1
From 85 benchmarks. Plenty of data behind it.
Not enough data yet$0.018$0.03131K
52.2
Fair value (52.2). Costs per task about what its benchmark results predict. Little data behind it, so treat it as a rough estimate.
113.9
Epoch Capabilities IndexSource: Epoch AI Benchmarking Hub API
Mistral Nemo
🇫🇷
Maker headquartered in FR, according to the company's own website. The citation is on the model page.
Open weights. The weights are publicly available, according to the provider's model list or a public Hugging Face repo.
fp16
Inference by Ionet
98.1
From 85 benchmarks. Plenty of data behind it.
Not enough data yet$0.0166cache read $0.0108$0.0216128K
52.2
Fair value (52.2). Costs per task about what its benchmark results predict. Little data behind it, so treat it as a rough estimate.
113.9
Epoch Capabilities IndexSource: Epoch AI Benchmarking Hub API
Mistral Nemo
🇫🇷
Maker headquartered in FR, according to the company's own website. The citation is on the model page.
Open weights. The weights are publicly available, according to the provider's model list or a public Hugging Face repo.
eu
Inference by Mistral AI
98.1
From 85 benchmarks. Plenty of data behind it.
Not enough data yet$0.15cache read $0.015$0.15131K
52.2
Fair value (52.2). Costs per task about what its benchmark results predict. Little data behind it, so treat it as a rough estimate.
113.9
Epoch Capabilities IndexSource: Epoch AI Benchmarking Hub API
Mistral Nemo
🇫🇷
Maker headquartered in FR, according to the company's own website. The citation is on the model page.
Open weights. The weights are publicly available, according to the provider's model list or a public Hugging Face repo.
fp8
Inference by Parasail
98.1
From 85 benchmarks. Plenty of data behind it.
Not enough data yet$0.03$0.03131K
52.2
Fair value (52.2). Costs per task about what its benchmark results predict. Little data behind it, so treat it as a rough estimate.
113.9
Epoch Capabilities IndexSource: Epoch AI Benchmarking Hub API
Qwen2.5 Coder 7B Instruct
🇨🇳
Maker headquartered in CN, according to the company's own website. The citation is on the model page.
98.0
From 59 benchmarks. Plenty of data behind it.
77 ±11stale
Coding IQ 76.6, 95% interval 66–87. A fair amount of data behind it. Uses the model's best published results. It has its own scale, separate from llminfo IQ. Its newest result is over a year old.
$0.20$0.20Not published by the provider
54.5
Good value (54.5). Cheaper per task than its benchmark results predict. Little data behind it, so treat it as a rough estimate.
Not measured
DeepSeek R1 Distill Llama 70B
🇨🇳
Maker headquartered in CN, according to the company's own website. The citation is on the model page.
97.6
From 6 benchmarks. A fair amount of data behind it.
Not enough data yet$0.10$0.40Not published by the providerNot enough data yetNot measured
DeepSeek R1 Distill Llama 70B
🇨🇳
Maker headquartered in CN, according to the company's own website. The citation is on the model page.
deepseek/deepseek-r1-distill-llama-70b:free · free
97.6
From 6 benchmarks. A fair amount of data behind it.
Not enough data yet$0.00$0.00Not published by the providerNot enough data yetNot measured
Phi-3 Mini 128K Instruct
🇺🇸
Maker headquartered in US, according to the company's own website. The citation is on the model page.
Newer version
Newer versionPhi-3.5 Mini 128K InstructWe compared the version names in our catalogue.
97.6
From 41 benchmarks. Plenty of data behind it.
Not enough data yet$0.10$0.10Not published by the provider
44.5
Poor value (44.5). Costs more per task than its benchmark results predict. Little data behind it, so treat it as a rough estimate.
Not measured
Command R+
🇺🇸
Maker headquartered in US, according to the company's own website. The citation is on the model page.
97.3
From 220 benchmarks. Plenty of data behind it.
≈ 76 ±11
Coding IQ 75.7, 95% interval 65–87. Little data behind it, so treat it as a rough estimate. Uses the model's best published results. It has its own scale, separate from llminfo IQ.
$0.50$1.50Not published by the providerNot enough data yetNot measured
Command R Plus
🇺🇸
Maker headquartered in US, according to the company's own website. The citation is on the model page.
96.9
From 196 benchmarks. Plenty of data behind it.
Not enough data yet$3.00$15.00Not published by the provider
21.9
Poor value (21.9). Costs more per task than its benchmark results predict. A fair amount of data behind it.
Not measured
Qwen-VL Max
🇨🇳
Maker headquartered in CN, according to the company's own website. The citation is on the model page.
96.9
From 56 benchmarks. Plenty of data behind it.
Not enough data yet$0.80$3.20Not published by the provider
49.8
Fair value (49.8). Costs per task about what its benchmark results predict. Little data behind it, so treat it as a rough estimate.
Not measured
Pixtral 12B
🇫🇷
Maker headquartered in FR, according to the company's own website. The citation is on the model page.
Open weights. The weights are publicly available, according to the provider's model list or a public Hugging Face repo.
96.6
From 81 benchmarks. Plenty of data behind it.
Not enough data yet$0.10$0.10Not published by the provider
54.4
Good value (54.4). Cheaper per task than its benchmark results predict. Little data behind it, so treat it as a rough estimate.
Not measured
Grok 2 Vision 1212
🇺🇸
Maker headquartered in US, according to the company's own website. The citation is on the model page.
1212
96.5
From 34 benchmarks. Plenty of data behind it.
Not enough data yet$2.00$10.00Not published by the provider
31.4
Poor value (31.4). Costs more per task than its benchmark results predict. Little data behind it, so treat it as a rough estimate.
Not measured
Llama 3.1 8B Instruct
🇺🇸
Maker headquartered in US, according to the company's own website. The citation is on the model page.
Open weights. The weights are publicly available, according to the provider's model list or a public Hugging Face repo.
Newer version
Newer versionLlama-3.3-8B-InstructWe compared the version names in our catalogue.
95.6
From 343 benchmarks. Plenty of data behind it.
70 ±9
Coding IQ 69.9, 95% interval 61–79. A fair amount of data behind it. Uses the model's best published results. It has its own scale, separate from llminfo IQ.
$0.02$0.04131K
39.5
Poor value (39.5). Costs more per task than its benchmark results predict. A fair amount of data behind it.
Not measured
Llama 3.1 8B Instruct
🇺🇸
Maker headquartered in US, according to the company's own website. The citation is on the model page.
Open weights. The weights are publicly available, according to the provider's model list or a public Hugging Face repo.
Newer version
Newer versionLlama-3.3-8B-InstructWe compared the version names in our catalogue.
us-east-1
Inference by Amazon Web Services
95.6
From 343 benchmarks. Plenty of data behind it.
70 ±9
Coding IQ 69.9, 95% interval 61–79. A fair amount of data behind it. Uses the model's best published results. It has its own scale, separate from llminfo IQ.
$0.22$0.22131K
39.5
Poor value (39.5). Costs more per task than its benchmark results predict. A fair amount of data behind it.
Not measured
Llama 3.1 8B Instruct
🇺🇸
Maker headquartered in US, according to the company's own website. The citation is on the model page.
Open weights. The weights are publicly available, according to the provider's model list or a public Hugging Face repo.
Newer version
Newer versionLlama-3.3-8B-InstructWe compared the version names in our catalogue.
bf16
Inference by Coreweave
95.6
From 343 benchmarks. Plenty of data behind it.
70 ±9
Coding IQ 69.9, 95% interval 61–79. A fair amount of data behind it. Uses the model's best published results. It has its own scale, separate from llminfo IQ.
$0.22$0.22131K
39.5
Poor value (39.5). Costs more per task than its benchmark results predict. A fair amount of data behind it.
Not measured
Llama 3.1 8B Instruct
🇺🇸
Maker headquartered in US, according to the company's own website. The citation is on the model page.
Open weights. The weights are publicly available, according to the provider's model list or a public Hugging Face repo.
Newer version
Newer versionLlama-3.3-8B-InstructWe compared the version names in our catalogue.
fp8
Inference by DeepInfra
95.6
From 343 benchmarks. Plenty of data behind it.
70 ±9
Coding IQ 69.9, 95% interval 61–79. A fair amount of data behind it. Uses the model's best published results. It has its own scale, separate from llminfo IQ.
$0.02$0.04131K
39.5
Poor value (39.5). Costs more per task than its benchmark results predict. A fair amount of data behind it.
Not measured
Llama 3.1 8B Instruct
🇺🇸
Maker headquartered in US, according to the company's own website. The citation is on the model page.
Open weights. The weights are publicly available, according to the provider's model list or a public Hugging Face repo.
Newer version
Newer versionLlama-3.3-8B-InstructWe compared the version names in our catalogue.
Inference by Groq
95.6
From 343 benchmarks. Plenty of data behind it.
70 ±9
Coding IQ 69.9, 95% interval 61–79. A fair amount of data behind it. Uses the model's best published results. It has its own scale, separate from llminfo IQ.
$0.05cache read $0.025$0.08131K
39.5
Poor value (39.5). Costs more per task than its benchmark results predict. A fair amount of data behind it.
Not measured
Phi 4 Mini
🇺🇸
Maker headquartered in US, according to the company's own website. The citation is on the model page.
Open weights. The weights are publicly available, according to the provider's model list or a public Hugging Face repo.
microsoft/phi-4-mini-instruct
95.6
From 41 benchmarks. Plenty of data behind it.
Not enough data yet$0.08$0.35Not published by the provider
40.9
Poor value (40.9). Costs more per task than its benchmark results predict. Little data behind it, so treat it as a rough estimate.
Not measured
Gemma 3 4B IT
🇺🇸
Maker headquartered in US, according to the company's own website. The citation is on the model page.
Open weights. The weights are publicly available, according to the provider's model list or a public Hugging Face repo.
google/gemma-3-4b-it:free · free
95.2
From 249 benchmarks. Plenty of data behind it.
67 ±19
Coding IQ 66.8, 95% interval 48–86. A fair amount of data behind it. Uses the model's best published results. It has its own scale, separate from llminfo IQ.
$0.00$0.00Not published by the provider
38.3
Poor value (38.3). Costs more per task than its benchmark results predict. Little data behind it, so treat it as a rough estimate.
Not measured
Gemma 3 4B IT
🇺🇸
Maker headquartered in US, according to the company's own website. The citation is on the model page.
Open weights. The weights are publicly available, according to the provider's model list or a public Hugging Face repo.
95.2
From 249 benchmarks. Plenty of data behind it.
67 ±19
Coding IQ 66.8, 95% interval 48–86. A fair amount of data behind it. Uses the model's best published results. It has its own scale, separate from llminfo IQ.
$0.04cache read $0.02$0.05131K
38.3
Poor value (38.3). Costs more per task than its benchmark results predict. Little data behind it, so treat it as a rough estimate.
Not measured
Gemma 3 4B IT
🇺🇸
Maker headquartered in US, according to the company's own website. The citation is on the model page.
Open weights. The weights are publicly available, according to the provider's model list or a public Hugging Face repo.
bf16
Inference by DeepInfra
95.2
From 249 benchmarks. Plenty of data behind it.
67 ±19
Coding IQ 66.8, 95% interval 48–86. A fair amount of data behind it. Uses the model's best published results. It has its own scale, separate from llminfo IQ.
$0.05$0.10131K
38.3
Poor value (38.3). Costs more per task than its benchmark results predict. Little data behind it, so treat it as a rough estimate.
Not measured
Gemma 3 4B IT
🇺🇸
Maker headquartered in US, according to the company's own website. The citation is on the model page.
Open weights. The weights are publicly available, according to the provider's model list or a public Hugging Face repo.
bf16/flex · bf16
Inference by DeepInfra
95.2
From 249 benchmarks. Plenty of data behind it.
67 ±19
Coding IQ 66.8, 95% interval 48–86. A fair amount of data behind it. Uses the model's best published results. It has its own scale, separate from llminfo IQ.
$0.04$0.08131K
38.3
Poor value (38.3). Costs more per task than its benchmark results predict. Little data behind it, so treat it as a rough estimate.
Not measured
Gemma 3 4B IT
🇺🇸
Maker headquartered in US, according to the company's own website. The citation is on the model page.
Open weights. The weights are publicly available, according to the provider's model list or a public Hugging Face repo.
bf16
Inference by Parasail
95.2
From 249 benchmarks. Plenty of data behind it.
67 ±19
Coding IQ 66.8, 95% interval 48–86. A fair amount of data behind it. Uses the model's best published results. It has its own scale, separate from llminfo IQ.
$0.04cache read $0.02$0.05131K
38.3
Poor value (38.3). Costs more per task than its benchmark results predict. Little data behind it, so treat it as a rough estimate.
Not measured
95.0
From 6 benchmarks. A fair amount of data behind it.
Not enough data yet$0.80$1.20Not published by the providerNot enough data yetNot measured
Mistral 7B Instruct
🇫🇷
Maker headquartered in FR, according to the company's own website. The citation is on the model page.
≈ 95
From 32 benchmarks. Little data behind it, so treat it as a rough estimate.
Not enough data yet$0.028$0.054Not published by the provider
60.8
Good value (60.8). Cheaper per task than its benchmark results predict. Little data behind it, so treat it as a rough estimate.
Not measured
Mistral 7B Instruct
🇫🇷
Maker headquartered in FR, according to the company's own website. The citation is on the model page.
mistralai/mistral-7b-instruct:free · free
≈ 95
From 32 benchmarks. Little data behind it, so treat it as a rough estimate.
Not enough data yet$0.00$0.00Not published by the provider
60.8
Good value (60.8). Cheaper per task than its benchmark results predict. Little data behind it, so treat it as a rough estimate.
Not measured
Llama 2 13B Chat
🇺🇸
Maker headquartered in US, according to the company's own website. The citation is on the model page.
94.9
From 45 benchmarks. A fair amount of data behind it.
Not enough data yet$0.22$0.22Not published by the providerNot enough data yetNot measured
NousResearch: Hermes 2 Pro - Llama-3 8B
🇺🇸
Maker headquartered in US, according to the company's own website. The citation is on the model page.
Open weights. The weights are publicly available, according to the provider's model list or a public Hugging Face repo.
nousresearch/hermes-2-pro-llama-3-8b
94.7
From 6 benchmarks. A fair amount of data behind it.
Not enough data yet$0.025$0.04Not published by the providerNot enough data yetNot measured
DeepSeek R1 Distill Llama 8B
🇺🇸
Maker headquartered in US, according to the company's own website. The citation is on the model page.
94.4
From 8 benchmarks. A fair amount of data behind it.
Not enough data yet$0.04$0.04Not published by the providerNot enough data yetNot measured
Molmo 7B-D
🇺🇸
Maker headquartered in US, according to the company's own website. The citation is on the model page.
93.4
From 44 benchmarks. Plenty of data behind it.
Not enough data yet$0.10$0.20Not published by the providerNot enough data yetNot measured

Data credits: Epoch AI Benchmarking Hub, CC BY 4.0 (licence)

How this data works