Spaces:
Sleeping
Sleeping
| { | |
| "qwen2-7b-instruct": { | |
| "feature": "qwen2-7b-instruct ‒ Chinese & English LLM for language, coding, mathematics and reasoning; costs $0.20 per M input tokens and $0.20 per M output tokens on Together AI.", | |
| "input_price": 0.20, | |
| "output_price": 0.20, | |
| "model": "qwen/qwen2-7b-instruct" | |
| }, | |
| "qwen2.5-7b-instruct": { | |
| "feature": "qwen2.5-7b-instruct ‒ upgraded Qwen with stronger multilingual capability; priced at $0.30 /M input and $0.30 /M output.", | |
| "input_price": 0.20, | |
| "output_price": 0.20, | |
| "model": "qwen/qwen2.5-7b-instruct" | |
| }, | |
| "gemma-7b": { | |
| "feature": "gemma-7b ‒ Google’s lightweight 7 B model for text and code; Together cost is $0.20 /M input and $0.20 /M output.", | |
| "input_price": 0.20, | |
| "output_price": 0.20, | |
| "model": "google/gemma-7b" | |
| }, | |
| "codegemma-7b": { | |
| "feature": "codegemma-7b ‒ Gemma variant focused on code generation & completion; $0.20 /M input, $0.20 /M output.", | |
| "input_price": 0.20, | |
| "output_price": 0.20, | |
| "model": "google/codegemma-7b" | |
| }, | |
| "gemma-2-9b-it": { | |
| "feature": "gemma-2-9b-it ‒ 2.9 B instruction-tuned Gemma for general text; ultralow $0.10 /M input and $0.10 /M output.", | |
| "input_price": 0.10, | |
| "output_price": 0.10, | |
| "model": "google/gemma-2-9b-it" | |
| }, | |
| "llama-3.1-8b-instruct": { | |
| "feature": "llama-3.1-8b-instruct ‒ Meta’s 8 B Llama-3 series for chat & reasoning; $0.20 /M input and $0.20 /M output.", | |
| "input_price": 0.20, | |
| "output_price": 0.20, | |
| "model": "meta/llama-3.1-8b-instruct" | |
| }, | |
| "granite-3.0-8b-instruct": { | |
| "feature": "granite-3.0-8b-instruct ‒ IBM small LLM supporting RAG, summarization & code; $0.20 /M input, $0.20 /M output.", | |
| "input_price": 0.20, | |
| "output_price": 0.20, | |
| "model": "ibm/granite-3.0-8b-instruct" | |
| }, | |
| "llama3-chatqa-1.5-8b": { | |
| "feature": "llama3-chatqa-1.5-8b ‒ NVIDIA fine-tuned 8 B for QA & reasoning; $0.20 /M input and output.", | |
| "input_price": 0.20, | |
| "output_price": 0.20, | |
| "model": "nvidia/llama3-chatqa-1.5-8b" | |
| }, | |
| "mistral-nemo-12b-instruct": { | |
| "feature": "mistral-nemo-12b-instruct ‒ 12 B model combining Mistral and NeMo tech; $0.30 /M input, $0.30 /M output.", | |
| "input_price": 0.30, | |
| "output_price": 0.30, | |
| "model": "nv-mistralai/mistral-nemo-12b-instruct" | |
| }, | |
| "mistral-7b-instruct-v0.3": { | |
| "feature": "mistral-7b-instruct-v0.3 ‒ fast 7 B model for instruction following; $0.20 /M in & out.", | |
| "input_price": 0.20, | |
| "output_price": 0.20, | |
| "model": "mistralai/mistral-7b-instruct-v0.3" | |
| }, | |
| "llama-3.3-nemotron-super-49b-v1": { | |
| "feature": "llama-3.3-nemotron-super-49b-v1 ‒ 49 B Nemotron with high accuracy; $0.90 /M input and output.", | |
| "input_price": 0.90, | |
| "output_price": 0.90, | |
| "model": "nvidia/llama-3.3-nemotron-super-49b-v1" | |
| }, | |
| "llama-3.1-nemotron-51b-instruct": { | |
| "feature": "llama-3.1-nemotron-51b-instruct ‒ 51 B NVIDIA alignment model; $0.90 /M in & out.", | |
| "input_price": 0.90, | |
| "output_price": 0.90, | |
| "model": "nvidia/llama-3.1-nemotron-51b-instruct" | |
| }, | |
| "llama3-chatqa-1.5-70b": { | |
| "feature": "llama3-chatqa-1.5-70b ‒ 70 B chat-optimized Llama; $0.90 /M input and output.", | |
| "input_price": 0.90, | |
| "output_price": 0.90, | |
| "model": "nvidia/llama3-chatqa-1.5-70b" | |
| }, | |
| "llama-3.1-70b-instruct": { | |
| "feature": "llama-3.1-70b-instruct ‒ Meta 70 B for complex conversations; $0.90 /M input/output.", | |
| "input_price": 0.90, | |
| "output_price": 0.90, | |
| "model": "meta/llama3-70b-instruct" | |
| }, | |
| "llama3-70b-instruct": { | |
| "feature": "llama3-70b-instruct ‒ alternate naming of Meta’s 70 B; $0.90 /M input & output.", | |
| "input_price": 0.90, | |
| "output_price": 0.90, | |
| "model": "meta/llama-3.1-8b-instruct" | |
| }, | |
| "granite-34b-code-instruct": { | |
| "feature": "granite-34b-code-instruct ‒ 34 B IBM coder model; $0.80 /M input and output.", | |
| "input_price": 0.80, | |
| "output_price": 0.80, | |
| "model": "ibm/granite-34b-code-instruct" | |
| }, | |
| "mixtral-8x7b-instruct-v0.1": { | |
| "feature": "mixtral-8×7b-instruct-v0.1 ‒ 56 B MoE (8×7 B) for creative text; $0.60 /M input/output.", | |
| "input_price": 0.60, | |
| "output_price": 0.60, | |
| "model": "mistralai/mixtral-8x7b-instruct-v0.1" | |
| }, | |
| "deepseek-r1": { | |
| "feature": "deepseek-r1 ‒ 671 B-param reasoning powerhouse; Together charges $3 /M input tokens and $7 /M output tokens.", | |
| "input_price": 0.55, | |
| "output_price": 2.19, | |
| "model": "deepseek-ai/deepseek-r1" | |
| }, | |
| "mixtral-8x22b-instruct-v0.1": { | |
| "feature": "mixtral-8×22b-instruct-v0.1 ‒ 176 B MoE (8×22 B); $1.20 /M input and output.", | |
| "input_price": 1.20, | |
| "output_price": 1.20, | |
| "model": "mistralai/mixtral-8x22b-instruct-v0.1" | |
| }, | |
| "palmyra-creative-122b": { | |
| "feature": "palmyra-creative-122b ‒ 122 B parameter model from Writer, optimized for creative and marketing content generation; $1.80 /M input and $1.80 /M output.", | |
| "input_price": 1.80, | |
| "output_price": 1.80, | |
| "model": "writer/palmyra-creative-122b" | |
| } | |
| } | |