o3 on Vidman AI
o3
Pricing
₹ per 1M tokens · as of 16 October 2026| Input | ₹192 |
| Output | ₹768 |
| Cached input | ₹48 |
Flat rate: every request bills the same. Rupee prices convert at the platform exchange rate billing uses (₹96/USD today). Full catalog and comparisons on the pricing comparison page.
About this model
OpenAI's well-rounded reasoning model for math, science, coding and visual reasoning. Succeeded by GPT-5.
Not sure? Vidman AI Adaptive routes to the best model per request, including this one.
Model specifications
API data- Modality
- Multimodal
- Context
- 2,00,000 tokens
- Maximum output
- 1,00,000 tokens
- Reasoning
- Always
- Effort levels
- low, medium, high
- Default effort
- medium
- Caching
- Automatic
- Cache minimum
- None
- Cache TTL
- Provider managed
- Route
- Vidman AI managed
From the serverless catalog, fetched 16 October 2026.
Measured performance
API observations over 7d; not an SLA.
- TTFT p50
- 1,500 ms
- TTFT p95
- 1,500 ms
- Throughput p50
- 1 tok/s
- Throughput average
- 0 tok/s
curl https://api.vidman.ai/v1/chat/completions \
-H "x-api-key: $VIDMAN_API_KEY" \
-H "Content-Type: application/json" \
-d '{
"model": "o3",
"messages": [{ "role": "user", "content": "Hello" }]
}'Existing OpenAI SDK code works by changing the base URL to https://api.vidman.ai/v1 and the key.
Model card
o3 is a well-rounded and powerful reasoning model across domains. OpenAI describes it as setting a new standard for math, science, coding and visual reasoning tasks and as excelling at technical writing and instruction-following. Use it to think through multi-step problems that involve analysis across text, code and images.
o3 is succeeded by GPT-5.
Model details
| Property | Value |
|---|---|
| Developer | OpenAI |
| Model ID | o3 |
| OpenAI default snapshot | o3-2025-04-16 |
| Input | Text, image |
| Output | Text |
| Context window | 200,000 tokens |
| Max output tokens | 100,000 |
| Knowledge cutoff | Jun 01, 2024 |
| Reasoning | Yes |
*Details above are from OpenAI's official model page.