DeepSeek route
deepseek-v4-flash-vision-exp
DeepSeek V4 Flash Vision API access.
Experimental vision sibling of DeepSeek V4 Flash. RouterPlex bills DeepSeek's off-peak $0.22 / $0.66 per 1M at all hours; the lab's peak window is double that. The lab may change or withdraw it. See the DeepSeek V4 Flash Vision API pricing notes.
Input / 1M tokens
$0.22
Output / 1M tokens
$0.66
Cached input / 1M tokens
$0.007
Context window
1M
Provider
DeepSeek
Workload calculator
Put the token rate into context.
These examples apply the published input and output rates directly. Actual cost follows the exact tokens your client sends and the model returns.
This route is the 13th cheapest of 64 comparable text models and cheaper than 51 on the same key. Its context window ranks 29th of 64.
Short request
$0.0029
10K input + 1K output
Agent turn
$0.029
100K input + 10K output
Exact token billing
No request minimum
One model ID change
OpenAI compatible
Cost neighbourhood
Nearby routes, same workload.
Every comparison uses the same 100K-input / 10K-output turn. Changing route is a one-line model ID edit.
Price does not measure task quality. For generation routes, see GPT Image 2.
First request
Route it in one call.
Use any OpenAI SDK or tool. Keep the request shape, point the base URL at RouterPlex, and set the model ID shown here.
Read the quickstartcurl https://api.routerplex.com/v1/chat/completions \-H "Authorization: Bearer $ROUTERPLEX_API_KEY" \-H "Content-Type: application/json" \-d '{"model": "deepseek-v4-flash-vision-exp","messages": [{"role": "user", "content": "Hello!"}]}'
Route notes
Questions before production traffic.
01How much does the DeepSeek V4 Flash Vision API cost?+
Through RouterPlex, DeepSeek V4 Flash Vision costs $0.22 per 1M input tokens and $0.66 per 1M output tokens — the vendor list price with 0% markup, billed per token from a prepaid balance you top up from $5. Cached input tokens are billed at $0.007 per 1M instead of the full input rate, which is most of the prompt on a repeated agent or coding loop. DeepSeek has retired this preview: it now serves the id with DeepSeek-V4.1-Flash and bills it at the Flash price ($0.15 input / $0.60 output per 1M tokens off-peak, double during peak hours). RouterPlex serves it through a separate upstream and bills a flat $0.22 input / $0.66 output per 1M tokens at all hours. For the same model at DeepSeek's time-weighted rate, use deepseek-v4.1-flash.
02What does one DeepSeek V4 Flash Vision agent turn cost?+
A 100K-token input with a 10K-token reply costs about $0.029 on DeepSeek V4 Flash Vision. That makes it the 13th cheapest of 64 comparable text models on RouterPlex, roughly 2.4× the cost of the cheapest option (Muse Spark 1.2 Contributor).
03What is a cheaper alternative to DeepSeek V4 Flash Vision?+
Hy3 from Tencent Hunyuan is the closest cheaper model on RouterPlex — $0.028 per agent turn against $0.029 for DeepSeek V4 Flash Vision, at $0.20/1M input and $0.80/1M output. Both run on the same key and the same endpoint, so switching is a one-line model ID change.
04Is DeepSeek V4 Flash Vision cheaper than Step 3.7 Flash?+
Yes. Per agent turn, DeepSeek V4 Flash Vision costs about $0.029 against $0.032 for Step 3.7 Flash — 9% less.
05What is the context window of DeepSeek V4 Flash Vision?+
DeepSeek V4 Flash Vision supports a 1M token context window on RouterPlex, the 29th largest of the 64 models that publish one.
06Can I use DeepSeek V4 Flash Vision with the OpenAI SDK?+
Yes. RouterPlex serves DeepSeek V4 Flash Vision through an OpenAI-compatible endpoint at https://api.routerplex.com/v1 — point any OpenAI SDK or tool at that base URL with your RouterPlex key and set model to "deepseek-v4-flash-vision-exp".
Provider family
More from DeepSeek.
Deploy this route
Put DeepSeek V4 Flash Vision behind one key.
Top up from $5 with no subscription required.