Together AI: Models Intelligence, Performance & Price
This analysis is intended to support you in choosing the best model provided by Together AI for your use-case.
Most Intelligent
Intelligence index
Total 16 models
Fastest
Output speed
Total 16 models
Lowest Price
Blended price (per 1M tokens)
Total 16 models
Together AI offers 16 models, each with different intelligence, performance, and pricing characteristics. Below is a comparison of the key metrics across models.
- For intelligence, the top models on Together AI are Kimi K3 (max) (60), Qwen3.8 2.4T A95B (58), and GLM-5.3-Flash (57).
- For output speed, the fastest models are Kimi K2.7 Code (258 t/s), Qwen3.8 2.4T A95B (157 t/s), and GLM-5.2 (max) (145 t/s). Speed varies significantly across models, with a 92% difference between the fastest and slowest.
- For latency, Qwen3.5 9B (Non-reasoning) (FP8) (0.80s), Gemma 4 31B (Non-reasoning) (FP8) (1.21s), and Llama 3.3 70B Turbo (1.61s) offer the lowest time to first answer token.
- For pricing, GLM-5.3-Flash ($0.10), Qwen3.5 9B (FP8) ($0.18), and Qwen3.5 9B (Non-reasoning) (FP8) ($0.18) offer the lowest blended prices per 1M tokens.
- For context window size, Kimi K3 (max) (1M), GLM-5.3-Flash (1M), and MiniMax-M3 (1M) support the largest context windows on Together AI.
Highlights
Intelligence Evaluations
Artificial Analysis Intelligence Index
Intelligence Evaluations
Agentic real-world work tasks, (Elo-500)/2000
Agentic tool use
Agentic coding & terminal use
Coding
Reasoning & knowledge
Scientific reasoning
Physics reasoning
Knowledge
1 - hallucination rate
Long context reasoning
Agentic knowledge work, Elo
Agentic SaaS workflows
Legal agentic work, criterion pass rate
Agentic business operations
Quantitative analysis on spreadsheets & documents
Instruction following
Long-horizon agentic tasks
Kubernetes incident root-cause analysis
Visual reasoning
Intelligence Index vs. Price
Context Window
Context Window
Pricing
Intelligence Index vs. Price
Performance Summary
Output Speed vs. Price
Speed
Measured by Output Speed (tokens per second)
Output Speed
Latency
Measured by Time (seconds) to First Token
Latency: Time To First Answer Token
End-to-End Response Time
Seconds to output 500 tokens, calculated based on time to first token, 'thinking' time for reasoning models, and output speed
End-to-End Response Time vs. Price
Further Analysis | |||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|
Kimi K3 (max) | 1.05M | Open | 60 | $1.57 | 53 | 1.43 | 48.78 | 37.88 | |||
Qwen3.8 2.4T A95B | 262k | Open | 58 | $1.47 | 157 | 1.13 | 17.04 | 12.72 | |||
GLM-5.3-Flash | 1.05M | Open | 57 | $0.12 | 87 | 0.59 | 29.40 | 23.05 | |||
GLM-5.2 (max) | 262k | Open | 53 | $0.47 | 145 | 0.83 | 18.03 | 13.76 | |||
MiniMax-M3 | 1M | Open | 45 | $0.18 | 89 | 0.75 | 28.91 | 22.53 | |||
Kimi K2.7 Code | 262k | Open | 43 | $0.27 | 258 | 0.43 | 10.99 | 8.62 | |||
Inkling | 524k | Open | 42 | $0.48 | 135 | 0.83 | 19.30 | 14.78 | |||
Nemotron 3 Ultra | 512k | Open | 38 | $0.40 | 135 | 0.73 | 21.33 | 16.89 | |||
 Muse Glimmer (high) | 131k | Open | 35 | $0.08 | 101 | 0.85 | 25.51 | 19.73 | |||
MiniMax-M2.5 (FP4) | 197k | Open | 34* | -- | -- | -- | -- | -- | |||
GLM-4.7 | 203k | Open | 34 | $0.28 | -- | -- | -- | -- | |||
 Gemma 4 31B (FP8) | 262k | Open | 30 | $0.10 | 52 | 1.35 | 44.67 | 33.64 | |||
GLM-4.6 | 203k | Open | 29 | $0.33 | -- | -- | -- | -- | |||
GLM-4.7 (Non-reasoning) | 203k | Open | 27* | -- | -- | -- | -- | -- | |||
gpt-oss-120b (high) | 131k | Open | 24 | $0.07 | 80 | 0.62 | 31.83 | 24.96 | |||
GLM-4.6 (Non-reasoning) | 203k | Open | 23* | -- | -- | -- | -- | -- | |||
Gemma 4 31B (Non-reasoning) (FP8) | 262k | Open | 22 | $0.11 | 43 | 1.21 | 12.95 | -- | |||
Qwen3.5 9B (FP8) | 262k | Open | 22 | $0.30 | 93 | 0.78 | 27.52 | 21.39 | |||
 Apriel-v1.5-15B-Thinker | 131k | Open | 22* | -- | -- | -- | -- | -- | |||
 DeepSeek V3.1 (Non-reasoning) | 131k | Open | 21* | -- | -- | -- | -- | -- | |||
Qwen3 Coder Next (FP8) | 262k | Open | 21 | $0.47 | -- | -- | -- | -- | |||
Apriel-v1.6-15B-Thinker | 131k | Open | 21* | -- | -- | -- | -- | -- | |||
Qwen3.5 9B (Non-reasoning) (FP8) | 262k | Open | 21* | -- | 94 | 0.80 | 6.13 | -- | |||
DeepSeek R1 0528 | 164k | Open | 20* | -- | -- | -- | -- | -- | |||
DeepSeek R1 (Jan) | 164k | Open | 19 | $0.50 | -- | -- | -- | -- | |||
GLM-4.5-Air (FP8) | 131k | Open | 17* | -- | -- | -- | -- | -- | |||
gpt-oss-120b (low) | 131k | Open | 15 | $0.02 | 90 | 0.73 | 28.59 | 22.28 | |||
Llama 4 Maverick | 1.05M | Open | 14 | $0.04 | -- | -- | -- | -- | |||
Llama 3.3 70B Turbo | 131k | Open | 9* | -- | 52 | 1.61 | 11.24 | -- | |||
Gemma 3n E4B | 32.8k | Open | 1* | -- | -- | -- | -- | -- | |||
 Cogito v2.1 | 164k | Open | -- | -- | -- | -- | -- | -- | |||
Key definitions
Frequently Asked Questions
Common questions about Together AI
Together AI offers 16 models that we track: Kimi K3 (max), Qwen3.8 2.4T A95B, GLM-5.3-Flash, GLM-5.2 (max), MiniMax-M3, Kimi K2.7 Code, Inkling, Nemotron 3 Ultra, Muse Glimmer (high), Gemma 4 31B (FP8), gpt-oss-120b (high), Gemma 4 31B (Non-reasoning) (FP8), Qwen3.5 9B (FP8), Qwen3.5 9B (Non-reasoning) (FP8), gpt-oss-120b (low), and Llama 3.3 70B Turbo.
The most intelligent model available on Together AI is Kimi K3 (max) with an Intelligence Index score of 60.
The fastest model on Together AI by output speed is Kimi K2.7 Code at 258.5 tokens per second.
The model with the lowest time to first answer token on Together AI is Qwen3.5 9B (Non-reasoning) (FP8) at 0.80s. Lower latency means faster initial response time.
The most affordable model on Together AI by blended price is GLM-5.3-Flash at $0.10 per 1M tokens (7:2:1 cache hit/input/output ratio).
Prices on Together AI vary up to 23x across models, from $0.10 per 1M tokens for GLM-5.3-Flash to $2.31 per 1M tokens for Kimi K3 (max).
Yes, Together AI offers an OpenAI-compatible API, making it easy to switch from OpenAI or use existing OpenAI SDK integrations.
Yes, all 16 models on Together AI support JSON mode for structured output.
Yes, all 16 models on Together AI support function calling (tool use).
Yes, Together AI offers 13 reasoning models: Kimi K3 (max), Qwen3.8 2.4T A95B, GLM-5.3-Flash, GLM-5.2 (max), MiniMax-M3, Kimi K2.7 Code, Inkling, Nemotron 3 Ultra, Muse Glimmer (high), Gemma 4 31B (FP8), gpt-oss-120b (high), Qwen3.5 9B (FP8), and gpt-oss-120b (low). Reasoning models use extended thinking to work through complex problems before providing an answer.
Yes, all 16 models on Together AI are open weight models.
Yes, provider performance can vary over time due to infrastructure changes, load balancing, and updates. We continuously benchmark all providers and display historical performance trends in the "Over Time" charts.
When choosing a model on Together AI, consider: intelligence (for quality-sensitive tasks), output speed (for throughput-intensive tasks), latency (for interactive applications requiring quick first responses), pricing (for cost-sensitive workloads), and features like context window size, JSON mode, or function calling support.