Phi-4
by MicrosoftChat
Open-weight instruction model for adaptable chat and self-hosted production workloads
- Context
- 16.4K tokens
- Max output
- 16,384 tokens
- Parameters
- 14.7B total
- Released
- 2025-01-10
- Weights
- Open
- Licence
- mit
microsoft/phi-4
Specifications
Full facts as recorded, incl. what is unknown.
- Released
- 2025-01-10
- Weights
- Opennot gated
- License
- mit
- Input
- Text
- Output
- Text
- Context window
- 16,384 max across endpoints
- Max output
- 16,384
- Providers
- 3 + 2 gateways, not counted as providers
- Epoch Capabilities Index
- not evaluated
- Direct price range
- $0.070–$0.125 in · $0.14–$0.50 out per 1M tokens · across all 3 providers
More details
- Last updated
- 2025-01-10
- Knowledge cutoff
- 2023-10
- Weights link
- Hugging Face · microsoft/phi-4
- Parameters
- 14.7B total
Capabilities
Across all endpoints: supported if any endpoint supports it.
- Tool calling
- Parallel toolsunknown
- Structured output
- Reasoning
- Image input
- PDF input
- Audio input
- Audio outputunknown
- Video input
- Prompt cachingunknown
- Web searchunknown
- Computer useunknown
- System messagesunknown
- Assistant prefillunknown
- Temperature
- Streamingunknown
- Attachments
Endpoints
One row per provider and variant; gateways are listed separately.
Loading endpoints…
Price history
Atom feedNo change since 19 Oct 2026. The cheapest direct price has been $0.07 in / $0.14 out per 1M tokens at Deep Infra on every data refresh since then.
Tracked from 19 Oct 2026. Headline-eligible direct prices only, so a price the sanity check flags never draws a spike. Series as JSON
Cost calculator
Loading endpoints…
Similarly priced models
Blended rate per 1M tokens, each at its own cheapest provider (10M input : 2M output mix).
| Model | Blended/1M | In/1M | Out/1M | Context | vs Phi-4 |
|---|---|---|---|---|---|
| Hy-MT2-1.8BTencent | $0.0662 | $0.044 | $0.177 | 8.2K | +1% |
| Tencent Hy-MT2-LiteTencent | $0.0662 | $0.044 | $0.177 | 8K | +1% |
| Qwen 3.5 9B (Q4_K_M)Alibaba (Qwen) | $0.0667 | $0.05 | $0.15 | 262K | +2% |
| DiffusionGemmaGoogle | $0.0667 | $0.05 | $0.15 | 262K | +2% |
| Nemotron 3.5 Content SafetyNVIDIA | $0.0667 | $0.05 | $0.15 | 128K | +2% |
| GLM-5.3-FlashZ.ai (Zhipu AI) | $0.0667 | $0.048 | $0.16 | 1.05M | +2% |
| Seed 1.6 FlashByteDance Seed | $0.0617 | $0.03 | $0.22 | 256K | −6% cheaper |
| Laguna XS 2.1Poolside | $0.07 | $0.06 | $0.12 | 262K | +7% |
More from Microsoft
Newest Microsoft models in the catalogue. All Microsoft models
| Model | Released | Context | Listed price |
|---|---|---|---|
| Microsoft-Decision-1 | 2026-10-09 | 32.8K | unpriced |
| MAI-Voice-2.1 | 2026-10-01 | — | unpriced |
| MAI-Voice-2.1-Flash | 2026-10-01 | — | unpriced |
| MAI-Transcribe-2-Streaming | 2026-10-01 | — | unpriced |
| MAI-Image-2.6 | 2026-09-04 | 4.1K | unpriced |
| MAI-Image-2.6-Flash | 2026-09-04 | 4.1K | unpriced |
| MAI-Transcribe 2 | 2026-09-03 | — | unpriced |
| MAI-Code-1.1-Flash | 2026-08-11 | 256K | $0.20 via GitHub Copilot |