# DeepSeek V4.1 Flash API

Fast 1M-context reasoning model with aggressive cache pricing, built for high-volume agent traffic. Always the latest version, now 1017.

Model id: `deepseek/deepseek-v4.1-flash`. Page: https://openscale.so/models/deepseek-v4.1-flash

## Price per 1M tokens (USD)

| Mode | Input | Cached input | Output |
| --- | --- | --- | --- |
| Standard | $0.06 | $0.003 | $0.24 |
| Fast mode | $0.28 | $0.028 | $1.10 |

50% off the list price of $0.12 input, $0.006 cached input and $0.48 output.

## Specs

- Context length: 1,048,568 tokens
- Max output: 384,000 tokens per request
- Output speed: 200+ tokens per second
- Rate limits: 100 requests and 10,000,000 tokens per minute per API key
- Text and image input, text output.
- Tool calling.
- Structured outputs with a JSON schema, and JSON mode.
- Reasoning you can turn on or off; the reasoning is returned as reasoning. Reasoning effort: low, high, max.

## Versions

| Model id | Version | Released | Behavior |
| --- | --- | --- | --- |
| `deepseek/deepseek-v4.1-flash` | 1017 | 17 Oct 2026 | Moves to each new tested version |
| `deepseek/deepseek-v4.1-flash-1017` | 1017 | 17 Oct 2026 | Never changes |
| `deepseek/deepseek-v4.1-flash-0910` | 0910 | 10 Sep 2026 | Never changes |
| `deepseek/deepseek-v4.1-flash:fast` | 1017 | 17 Oct 2026 | Moves to each new tested version, fast mode |
| `deepseek/deepseek-v4.1-flash-1017:fast` | 1017 | 17 Oct 2026 | Never changes, fast mode |
| `deepseek/deepseek-v4.1-flash-0910:fast` | 0910 | 10 Sep 2026 | Never changes, fast mode |

## Quick start

### curl

```bash
curl https://api.openscale.so/v1/chat/completions \
  -H "Authorization: Bearer $OPENSCALE_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "model": "deepseek/deepseek-v4.1-flash",
    "messages": [{"role": "user", "content": "Explain what a token is in one sentence."}]
  }'
```

### OpenAI SDK (Python)

```python
import os
from openai import OpenAI

client = OpenAI(
    base_url="https://api.openscale.so/v1",
    api_key=os.environ["OPENSCALE_API_KEY"],
)

response = client.chat.completions.create(
    model="deepseek/deepseek-v4.1-flash",
    messages=[{"role": "user", "content": "Explain what a token is in one sentence."}],
)
print(response.choices[0].message.content)
```

### Anthropic SDK (Python)

```python
import os
from anthropic import Anthropic

client = Anthropic(
    base_url="https://api.openscale.so/anthropic",
    api_key=os.environ["OPENSCALE_API_KEY"],
)

message = client.messages.create(
    model="deepseek/deepseek-v4.1-flash",
    max_tokens=1024,
    messages=[{"role": "user", "content": "Explain what a token is in one sentence."}],
)
print(next(block.text for block in message.content if block.type == "text"))
```

## Set up a coding agent

- [Claude Code](https://openscale.so/setup/claude-code.md)
- [Codex](https://openscale.so/setup/codex.md)
- [OpenCode](https://openscale.so/setup/opencode.md)
- [OpenClaw](https://openscale.so/setup/openclaw.md)
- [Cline](https://openscale.so/setup/cline.md)
- [GitHub Copilot](https://openscale.so/setup/copilot.md)
- [DeepSeek Harness](https://openscale.so/setup/deepseek-harness.md)
- [Hermes](https://openscale.so/setup/hermes.md)
- [Kilo Code](https://openscale.so/setup/kilo-code.md)
- [Pi](https://openscale.so/setup/pi.md)
- [Zed](https://openscale.so/setup/zed.md)
- [Factory Droid](https://openscale.so/setup/droid.md)

API reference: https://openscale.so/docs. Live model catalog: https://api.openscale.so/v1/models.

## FAQ

### How much does the DeepSeek V4.1 Flash API cost?

$0.06 per 1M input tokens, $0.003 per 1M cached input tokens and $0.24 per 1M output tokens. That is 50% off the list price of $0.12, $0.006 and $0.48. You pay only for the tokens you use, from prepaid credit.

### Is the DeepSeek V4.1 Flash API OpenAI compatible?

Yes. Point any OpenAI SDK at https://api.openscale.so/v1 and use the model id deepseek/deepseek-v4.1-flash, with Chat Completions or Responses. Anthropic SDKs work too, at https://api.openscale.so/anthropic.

### What is the context length of DeepSeek V4.1 Flash?

1,048,568 tokens in total, input and output together, with up to 384,000 tokens of output per request.

### How fast is DeepSeek V4.1 Flash on OpenScale?

200+ tokens per second of output, a floor under the measured median. Fast mode (deepseek/deepseek-v4.1-flash:fast) is ~2x standard, at $0.28 input and $1.10 output per 1M tokens.

### Which DeepSeek V4.1 Flash version does the API serve?

deepseek/deepseek-v4.1-flash always serves the latest version we have tested, now 1017. To stay on one version, call a pinned id: deepseek/deepseek-v4.1-flash-1017, deepseek/deepseek-v4.1-flash-0910. Pinned ids never change, and every version has the same price.

### Can I use DeepSeek V4.1 Flash in a coding agent?

Yes. OpenScale has a tested setup for 12 coding agents, including Claude Code, Codex, OpenCode, OpenClaw. One API key works for every agent.

### What are the rate limits for DeepSeek V4.1 Flash?

100 requests and 10,000,000 tokens per minute for each API key. Need more? Email hello@openscale.so.

### How do I start?

Create an account, add credit and create an API key. New accounts get free credit to try it.