DigitalOcean: Models Intelligence, Performance & Price
This analysis is intended to support you in choosing the best model provided by DigitalOcean for your use-case.
Most Intelligent
GLM-5.3 (max)
Kimi K3 (max)GLM-5.3-Flash
Qwen3.8 2.4T A95B
DeepSeek V4.1 Flash (max)Intelligence Index
Total 21 models
Fastest
Nemotron 3 Nano Omni 30B A3B
MiniMax-M2.5DeepSeek V4.1 Flash (non-reasoning)DeepSeek V4.1 Flash (max)Qwen3.8 2.4T A95BOutput speed
Total 21 models
Lowest Price
DeepSeek V4 Flash 0731 (max)
Ministral 3 14BGLM-5.3-Flash
Gemma 4 31BGemma 4 31B (non-reasoning)Blended price (per 1M tokens)
Total 21 models
DigitalOcean offers 21 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 DigitalOcean are GLM-5.3 (max) (45), Kimi K3 (max) (44), and GLM-5.3-Flash (42).
- For output speed, the fastest models are Nemotron 3 Nano Omni 30B A3B (228 t/s), MiniMax-M2.5 (83 t/s), and DeepSeek V4.1 Flash (non-reasoning) (78 t/s). Speed varies significantly across models, with a 252% difference between the fastest and slowest.
- For latency, DeepSeek V4.1 Flash (non-reasoning) (0.78s), DeepSeek V4 Pro 0813 (non-reasoning) (0.85s), and Llama 4 Maverick (0.99s) offer the lowest time to first answer token.
- For pricing, DeepSeek V4 Flash 0731 (max) ($0.08), Ministral 3 14B ($0.09), and GLM-5.3-Flash ($0.10) offer the lowest blended prices per 1M tokens.
- For context window size, GLM-5.3 (max) (1M), Kimi K3 (max) (1M), and GLM-5.3-Flash (1M) support the largest context windows on DigitalOcean.
Intelligence Evaluations
Artificial Analysis Intelligence Index
Intelligence Evaluations
Agentic knowledge work, (Elo-500)/2000
Knowledge
1 - hallucination rate
Agentic SaaS workflows
Agentic real-world work tasks, (Elo-500)/2000
Reasoning & knowledge
Professional document reasoning, All-pass
Agentic coding & terminal use
Long context reasoning
Coding
Physics reasoning
Agentic scientific research workflows in a terminal
Agentic software engineering in open-source repositories
Quantitative analysis on spreadsheets & documents
Legal agentic work, Hallucination-Gated All-Pass Rate
Agentic business operations
Kubernetes incident root-cause analysis
Medical long context reasoning
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
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 | 52 | 0.85 | 48.56 | 38.17 | |||
Kimi K3 (max) | 1.05M | Open | 44 | $1.90 | 52 | 1.10 | 49.00 | 38.32 | |||
GLM-5.3-Flash | 1.05M | Open | 42 | $0.42 | 50 | 0.85 | 51.16 | 40.25 | |||
Qwen3.8 2.4T A95B | 262k | Open | 40 | $1.81 | 65 | 0.89 | 39.45 | 30.85 | |||
DeepSeek V4.1 Flash (max) | 1M | Open | 39 | $1.75 | 77 | 0.80 | 33.08 | 25.82 | |||
DeepSeek V4 Pro 0813 (max) | 1.05M | Open | 36 | $1.61 | 55 | 0.85 | 46.27 | 36.33 | |||
DeepSeek V4 Flash 0731 (max) | 1.05M | Open | 34 | $0.16 | 47 | 0.90 | 54.30 | 42.72 | |||
GLM-5.3 (low) | 1.05M | Open | 34 | $0.84 | 55 | 0.91 | 45.99 | 36.06 | |||
Kimi K3 (low) | 1M | Open | 30 | $0.92 | 51 | 1.13 | 49.96 | 39.07 | |||
DeepSeek V4.1 Flash (non-reasoning) | 1.05M | Open | 25 | $1.20 | 78 | 0.78 | 7.23 | -- | |||
MiniMax-M2.5 | 197k | Open | 23* | -- | 83 | 0.99 | 31.11 | 24.10 | |||
DeepSeek V3.2 | 128k | Open | 21* | -- | 41 | 0.96 | 61.83 | 48.69 | |||
Qwen3.5 397B A17B (non-reasoning) | 131k | Open | 21* | -- | 15 | 1.50 | 35.56 | -- | |||
DeepSeek V4 Pro 0813 (non-reasoning) | 1.05M | Open | 20 | $0.43 | 53 | 0.85 | 10.34 | -- | |||
Qwen3.5 397B A17B | 131k | Open | 18 | $0.17 | 14 | 1.46 | 268.14 | 230.51 | |||
DeepSeek V3.2 (non-reasoning) | 128k | Open | 16* | -- | 42 | 1.08 | 12.90 | -- | |||
Gemma 4 31B | 262k | Open | 15 | $0.05 | 43 | 2.02 | 53.86 | 40.25 | |||
Gemma 4 31B (non-reasoning) | 262k | Open | 14* | -- | 45 | 1.94 | 13.08 | -- | |||
Nemotron 3 Nano Omni 30B A3B | 65.5k | Open | 10* | -- | 228 | 0.94 | 11.90 | 8.77 | |||
 Llama 4 Maverick | 128k | Open | 10* | -- | 14 | 0.99 | 36.63 | -- | |||
Ministral 3 14B | 262k | Open | 6 | $0.03 | 37 | 1.04 | 14.39 | -- | |||
Key definitions
Frequently Asked Questions
Common questions about DigitalOcean
DigitalOcean offers 21 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), DeepSeek V4 Pro 0813 (max), DeepSeek V4 Flash 0731 (max), GLM-5.3 (low), Kimi K3 (low), DeepSeek V4.1 Flash (non-reasoning), MiniMax-M2.5, DeepSeek V3.2, Qwen3.5 397B A17B (non-reasoning), DeepSeek V4 Pro 0813 (non-reasoning), Qwen3.5 397B A17B, DeepSeek V3.2 (non-reasoning), Gemma 4 31B, Gemma 4 31B (non-reasoning), Nemotron 3 Nano Omni 30B A3B, Llama 4 Maverick, and Ministral 3 14B.
The most intelligent model available on DigitalOcean is GLM-5.3 (max) with an Intelligence Index score of 45.
The fastest model on DigitalOcean by output speed is Nemotron 3 Nano Omni 30B A3B at 228.1 tokens per second.
The model with the lowest time to first answer token on DigitalOcean is DeepSeek V4.1 Flash (non-reasoning) at 0.78s. Lower latency means faster initial response time.
The most affordable model on DigitalOcean by blended price is DeepSeek V4 Flash 0731 (max) at $0.08 per 1M tokens (7:2:1 cache hit/input/output ratio).
Prices on DigitalOcean vary up to 31x across models, from $0.08 per 1M tokens for DeepSeek V4 Flash 0731 (max) to $2.31 per 1M tokens for Kimi K3 (low).
Yes, DigitalOcean offers an OpenAI-compatible API, making it easy to switch from OpenAI or use existing OpenAI SDK integrations.
Yes, all 21 models on DigitalOcean support JSON mode for structured output.
Yes, all 21 models on DigitalOcean support function calling (tool use).
Yes, DigitalOcean offers 14 reasoning models: GLM-5.3 (max), Kimi K3 (max), GLM-5.3-Flash, Qwen3.8 2.4T A95B, DeepSeek V4.1 Flash (max), DeepSeek V4 Pro 0813 (max), DeepSeek V4 Flash 0731 (max), GLM-5.3 (low), Kimi K3 (low), MiniMax-M2.5, DeepSeek V3.2, Qwen3.5 397B A17B, Gemma 4 31B, and Nemotron 3 Nano Omni 30B A3B. Reasoning models use extended thinking to work through complex problems before providing an answer.
Yes, all 21 models on DigitalOcean 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 DigitalOcean, 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.