Analyzes how recently AI models and their data have been updated. Track freshness scores, identify stale data, and see which providers keep their models most up to date.
100.0
out of 100
300
models updated
300
models updated
GPT-5.4 Pro
0m ago
Models grouped by how recently their data was last updated.
Models with the most recently updated data, ranked by freshness score.
| # | Model | Provider | Last Updated | Freshness |
|---|---|---|---|---|
| 1 | Llama Guard 3 8B | Meta | 0m ago | 100 |
| 2 | Qwen VL Plus | Alibaba | 0m ago | 100 |
| 3 | Aion-1.0 | aion-labs | 0m ago | 100 |
| 4 | Aion-1.0-Mini | aion-labs | 0m ago | 100 |
| 5 | Qwen VL Max | Alibaba | 0m ago | 100 |
| 6 | Qwen-Turbo | Alibaba | 0m ago | 100 |
| 7 | Qwen2.5 VL 72B Instruct | Alibaba | 0m ago | 100 |
| 8 | Qwen-Plus | Alibaba | 0m ago | 100 |
| 9 | Qwen-Max | Alibaba | 0m ago | 100 |
| 10 | Mistral Small 3 | Mistral AI | 0m ago | 100 |
| 11 | Sonar | Perplexity | 0m ago | 100 |
| 12 | MiniMax-01 | MiniMax | 0m ago | 100 |
| 13 | Nova Lite 1.0 | Amazon | 0m ago | 100 |
| 14 | Nova Micro 1.0 | Amazon | 0m ago | 100 |
| 15 | Gemma 3n 4B | 0m ago | 100 | |
| 16 | Spotlight | arcee-ai | 0m ago | 100 |
| 17 | Virtuoso Large | arcee-ai | 0m ago | 100 |
| 18 | Coder Large | arcee-ai | 0m ago | 100 |
| 19 | Llama Guard 4 12B | Meta | 0m ago | 100 |
| 20 | Qwen3 14B | Alibaba | 0m ago | 100 |
Models with the oldest data. These may have outdated scores or missing signal updates.
| # | Model | Provider | Last Updated | Freshness |
|---|---|---|---|---|
| 1 | GPT-5.4 Pro | OpenAI | 0m ago | 100 |
| 2 | GPT-5.4 | OpenAI | 0m ago | 100 |
| 3 | GPT-5.2 Pro | OpenAI | 0m ago | 100 |
| 4 | Claude Opus 4.6 (Fast) | Anthropic | 0m ago | 100 |
| 5 | Claude Opus 4.6 | Anthropic | 0m ago | 100 |
| 6 | GPT-5.2-Codex | OpenAI | 0m ago | 100 |
| 7 | GPT-5.2 | OpenAI | 0m ago | 100 |
| 8 | Grok 4.20 | xAI | 0m ago | 100 |
| 9 | GPT-5.3-Codex | OpenAI | 0m ago | 100 |
| 10 | GPT-5 Pro | OpenAI | 0m ago | 100 |
| 11 | Gemini 3 Flash Preview | 0m ago | 100 | |
| 12 | Grok 4 | xAI | 0m ago | 100 |
| 13 | Grok 4.20 Multi-Agent | xAI | 0m ago | 100 |
| 14 | GPT-5.1-Codex-Max | OpenAI | 0m ago | 100 |
| 15 | GPT-5 Codex | OpenAI | 0m ago | 100 |
| 16 | GPT-5 | OpenAI | 0m ago | 100 |
| 17 | GPT-5.3 Chat | OpenAI | 0m ago | 100 |
| 18 | GPT-5.1 | OpenAI | 0m ago | 100 |
| 19 | GPT-5.1-Codex | OpenAI | 0m ago | 100 |
| 20 | GPT-5.1-Codex-Mini | OpenAI | 0m ago | 100 |
How fresh is each provider's data? Sorted by average freshness score across all their models.
| Provider | Models | Avg Score |
|---|---|---|
| Perplexity | 5 | 100.0 |
| aion-labs | 3 | 100.0 |
| AI21 Labs | 1 | 100.0 |
| Amazon | 4 | 100.0 |
| rekaai | 2 | 100.0 |
| deepcogito | 1 | 100.0 |
| ByteDance | 5 | 100.0 |
| Upstage | 1 | 100.0 |
| Writer | 1 | 100.0 |
| essentialai | 1 | 100.0 |
| arcee-ai | 6 | 100.0 |
| Baidu | 7 | 100.0 |
| Liquid AI | 3 | 100.0 |
| NVIDIA | 9 | 100.0 |
| IBM | 2 | 100.0 |
| Kuaishou | 1 | 100.0 |
| Mistral AI | 18 | 100.0 |
| Alibaba | 49 | 100.0 |
| ~anthropic | 3 | 100.0 |
| inclusionai | 3 | 100.0 |
| Tencent | 2 | 100.0 |
| ~openai | 2 | 100.0 |
| 2 | 100.0 | |
| ~moonshotai | 1 | 100.0 |
| poolside | 2 | 100.0 |
| Meta | 9 | 100.0 |
| Xiaomi | 5 | 100.0 |
| Allen AI | 1 | 100.0 |
| Cohere | 3 | 100.0 |
| Moonshot AI | 5 | 100.0 |
| Microsoft | 2 | 100.0 |
| 23 | 100.0 | |
| MiniMax | 8 | 100.0 |
| StepFun | 1 | 100.0 |
| Cursor | 2 | 100.0 |
| Inception | 1 | 100.0 |
| Zhipu AI | 12 | 100.0 |
| DeepSeek | 12 | 100.0 |
| xAI | 11 | 100.0 |
| OpenAI | 57 | 100.0 |
| Anthropic | 14 | 100.0 |
Computed as 100 minus the number of hours since the last update, capped at 100 hours. A score of 100 means the model was just updated; a score of 0 means it has not been updated in over 100 hours (4+ days).
Models are grouped into five tiers based on update recency: Just Updated (<1h), Fresh (1-6h), Recent (6-24h), Aging (1-7d), and Stale (>7d). Models without a known update timestamp are classified as Stale.
Stale data can lead to inaccurate rankings and outdated signal scores. Models with fresh data reflect the latest pricing, capabilities, and benchmark results. Monitoring freshness helps identify data pipeline issues early.
Dive deeper into model tracking, release velocity, and weekly trend analysis across the AI ecosystem.
Data freshness measures how recently each AI model's information has been updated. A freshness score of 100 means the model was just updated, while 0 means its data is over 100 hours (4+ days) old.
Our data pipeline refreshes model information hourly from multiple API sources. Each update pulls the latest pricing, capabilities, and benchmark data for all tracked models.
Stale data can lead to inaccurate rankings and outdated signal scores. Models with fresh data reflect the latest pricing changes, capability updates, and benchmark results, ensuring you make decisions based on current information.
Models are grouped into five tiers: Just Updated (under 1 hour), Fresh (1-6 hours), Recent (6-24 hours), Aging (1-7 days), and Stale (over 7 days). Models without a known update timestamp are classified as Stale.