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
Updated
GLM-5.3 (max)
Kimi K3 (max)GLM-5.3-Flash
Qwen3.8 2.4T A95B
DeepSeek V4.1 Flash (max)Intelligence index
Total 14 models
Fastest
DeepSeek V4.1 Flash (max)GLM-5.2 (max)GLM-5.3 (max)
gpt-oss-120b (high)gpt-oss-120b (low)Output speed
Total 14 models
Lowest Price
GLM-5.3-FlashQwen3.5 9B (non-reasoning) (FP8)Qwen3.5 9B (FP8)DeepSeek V4.1 Flash (max)gpt-oss-120b (high)Blended price (per 1M tokens)
Total 14 models
Together AI offers 14 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 GLM-5.3 (max) (45), Kimi K3 (max) (44), and GLM-5.3-Flash (42).
- For output speed, the fastest models are DeepSeek V4.1 Flash (max) (290 t/s), GLM-5.2 (max) (274 t/s), and GLM-5.3 (max) (200 t/s). Speed varies significantly across models, with a 83% difference between the fastest and slowest.
- For latency, Qwen3.5 9B (non-reasoning) (FP8) (0.78s), Llama 3.3 70B Turbo (1.58s), and DeepSeek V4.1 Flash (max) (7.20s) offer the lowest time to first answer token.
- For pricing, GLM-5.3-Flash ($0.10), Qwen3.5 9B (non-reasoning) (FP8) ($0.18), and Qwen3.5 9B (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 DeepSeek V4.1 Flash (max) (1M) support the largest context windows on Together AI.
Intelligence Evaluations
Artificial Analysis Intelligence Index
Intelligence Evaluations
Agentic knowledge work, (Elo-500)/2000
Agentic real-world work tasks, (Elo-500)/2000
Agentic SaaS workflows
Agentic coding & terminal use
Coding
Reasoning & knowledge
Professional document reasoning, All-pass
Physics reasoning
Knowledge
1 - hallucination rate
Long context reasoning
Legal agentic work, criterion pass rate
Agentic business operations
Agentic scientific research workflows in a terminal
Quantitative analysis on spreadsheets & documents
Kubernetes incident root-cause analysis
Visual reasoning
Medical long context 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
Cache Behaviour
Cache Hit Rate
Cost per Task vs. Cache Hit Rate
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 | |||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|
GLM-5.3 (max) | 1.05M | Open | 45 | $1.97 | 212 | 0.56 | 12.36 | 9.44 | |||
Kimi K3 (max) | 1.05M | Open | 44 | $2.16 | 97 | 0.92 | 26.61 | 20.55 | |||
GLM-5.3-Flash | 1.05M | Open | 42 | $0.47 | 98 | 0.63 | 26.16 | 20.43 | |||
Qwen3.8 2.4T A95B | 262k | Open | 40 | $2.09 | 100 | 1.30 | 26.33 | 20.03 | |||
DeepSeek V4.1 Flash (max) | 1.05M | Open | 39 | $0.32 | 290 | 0.30 | 8.93 | 6.90 | |||
GLM-5.2 (max) | 262k | Open | 34 | $0.88 | 274 | 0.61 | 9.72 | 7.29 | |||
 MiniMax-M3 | 1M | Open | 29 | $0.58 | 45 | 1.30 | 56.51 | 44.16 | |||
 Inkling (xhigh) | 524k | Open | 25 | $0.91 | 129 | 0.65 | 20.02 | 15.50 | |||
MiniMax-M2.5 (FP4) | 197k | Open | 23* | -- | -- | -- | -- | -- | |||
GLM-4.7 | 203k | Open | 22* | -- | -- | -- | -- | -- | |||
GLM-4.6 | 203k | Open | 19* | -- | -- | -- | -- | -- | |||
 Muse Glimmer (high) | 131k | Open | 17 | $0.08 | 102 | 0.77 | 25.17 | 19.52 | |||
GLM-4.7 (non-reasoning) | 203k | Open | 17* | -- | -- | -- | -- | -- | |||
GLM-4.6 (non-reasoning) | 203k | Open | 15* | -- | -- | -- | -- | -- | |||
 Apriel-v1.5-15B-Thinker | 131k | Open | 14* | -- | -- | -- | -- | -- | |||
DeepSeek V3.1 (non-reasoning) | 131k | Open | 14* | -- | -- | -- | -- | -- | |||
Apriel-v1.6-15B-Thinker | 131k | Open | 13* | -- | -- | -- | -- | -- | |||
Qwen3.5 9B (non-reasoning) (FP8) | 262k | Open | 13* | -- | 93 | 0.77 | 6.14 | -- | |||
DeepSeek R1 0528 | 164k | Open | 13* | -- | -- | -- | -- | -- | |||
gpt-oss-120b (high) | 131k | Open | 12 | $0.11 | 166 | 0.42 | 15.46 | 12.03 | |||
DeepSeek R1 (Jan) | 164k | Open | 11 | $0.41 | -- | -- | -- | -- | |||
Qwen3.5 9B (FP8) | 262k | Open | 11 | $0.27 | 95 | 0.68 | 27.13 | 21.15 | |||
GLM-4.5-Air (FP8) | 131k | Open | 11* | -- | -- | -- | -- | -- | |||
gpt-oss-120b (low) | 131k | Open | 10* | -- | 158 | 0.38 | 16.20 | 12.65 | |||
Llama 4 Maverick | 1.05M | Open | 10* | -- | -- | -- | -- | -- | |||
Qwen3 Coder Next (FP8) | 262k | Open | 9 | $0.78 | -- | -- | -- | -- | |||
Llama 3.3 70B Turbo | 131k | Open | 8* | -- | 78 | 1.64 | 8.03 | -- | |||
 Cogito v2.1 | 164k | Open | -- | -- | -- | -- | -- | -- | |||
Key definitions
Frequently Asked Questions
Common questions about Together AI
Together AI offers 14 models that we track: GLM-5.3 (max), Kimi K3 (max), GLM-5.3-Flash, Qwen3.8 2.4T A95B, DeepSeek V4.1 Flash (max), GLM-5.2 (max), MiniMax-M3, Inkling (xhigh), Muse Glimmer (high), Qwen3.5 9B (non-reasoning) (FP8), gpt-oss-120b (high), Qwen3.5 9B (FP8), gpt-oss-120b (low), and Llama 3.3 70B Turbo.
The most intelligent model available on Together AI is GLM-5.3 (max) with an Intelligence Index score of 45.
The fastest model on Together AI by output speed is DeepSeek V4.1 Flash (max) at 289.8 tokens per second.
The model with the lowest time to first answer token on Together AI is Qwen3.5 9B (non-reasoning) (FP8) at 0.78s. 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 21x across models, from $0.10 per 1M tokens for GLM-5.3-Flash to $2.08 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 14 models on Together AI support JSON mode for structured output.
Yes, all 14 models on Together AI support function calling (tool use).
Yes, Together AI offers 12 reasoning models: GLM-5.3 (max), Kimi K3 (max), GLM-5.3-Flash, Qwen3.8 2.4T A95B, DeepSeek V4.1 Flash (max), GLM-5.2 (max), MiniMax-M3, Inkling (xhigh), Muse Glimmer (high), 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 14 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.