Token Harbor

The easiest way to access frontier AI models—with free models to get you started.
Come build with us ⚓

[Token Harbor @TokenHarborAI]

You really like DeepSeek V4.1 Flash. 👀
It’s already the #1 user-selected model on Token Harbor:
56% request share
75% token share
Nearly 1 in 2 active users tried it
And yes — it’s still free to use.

[Token Harbor @TokenHarborAI]

Kimi K2.8 👀

Which Pass should it land in?
Free? Agent? Office? Frontier?

I have a feeling I know what everyone's gonna say

[Lumina @LuminaBench]

🚨 Kimi K2.8 Preview just dropped in Kimi Code

From reading the docs it has performance close to K3 but apparently with "more efficient thinking"

• 1M context
• image + video input
• low / high / max reasoning

Kimi K3 is a very good model too, im hoping the "more efficient thinking" fixes the issues K3 had with being really slow, very surprised about this drop

[Token Harbor @TokenHarborAI]

DeepSeek V4.1 Flash keeps showing up

Now #6 on Design Arena — and still free on Token Harbor.

[Design Arena @DesignArena]

BREAKING: DeepSeek‑V4.1‑Flash takes 6th overall on Design Arena with an Elo of 1347!

This marks a 39-position jump over the next-highest DeepSeek model - and DeepSeek’s return to a top-10 placement on Design Arena.

The model ranks in the same performance band as Claude Fable 5.1 on real-world frontend design tasks.

Congratulations to the @deepseek_ai team!

[Token Harbor @TokenHarborAI]

Benchmark season is fun.

But “I ran both on my actual workload” is still my favorite benchmark.

[Artificial Analysis @ArtificialAnlys]

DeepSeek V4.1 Flash overtakes DeepSeek V4 Pro 0813 as DeepSeek’s new flagship model with a score of 40 on Artificial Analysis Intelligence Index. At just 552B parameters, it outperforms the Pro (1.6T) model while costing ~4x less per token, placing it just short of the Intelligence vs. Cost Pareto frontier because of its verbosity

@deepseek_ai has released DeepSeek V4.1 Flash, the successor to DeepSeek V4 Flash 0731. This is a 552B model features a new causal Encoder–Decoder architecture, allowing it to have just 8B active parameters for input and 16B active parameters for output.

On the first-party API, DeepSeek V4.1 Flash is priced at $0.30 per 1M input tokens and $1.20 per 1M output tokens, with cached input tokens priced at just $0.006 per 1M tokens, a 98% discount. Off-peak pricing provides a further 50% discount across input, cached input, and output tokens. V4.1 Flash is ~20% cheaper than DeepSeek V4 Flash 0731 and ~4x cheaper than DeepSeek V4 Pro 0813 while delivering a higher performance.

Key results:

➤ DeepSeek V4.1 Flash makes gains in agentic capabilities and long context reasoning. DeepSeek V4.1 Flash scores 27% in Terminal-Bench v4.0, more than double DeepSeek V4.0 Flash’s 27%. Its GDPval-AA v2 rises 164 Elo points from 1468 to 1632, placing V4.1 Flash ahead of Kimi K3 (1584). DeepSeek V4.1 Flash also scores 84% on AA-LCR v1.1, at the same level as GPT-5.6 Sol (84%) and Gemini 3.8 Flash (84%). With a 1M token context window, this makes it a strong option for long-document and large-repository work at flash-tier pricing.

➤ DeepSeek V4.1 Flash takes first place on AutomationBench-AA with 69%, equal to GPT-6 Astra (69%) and slightly above Grok 4.6 (67%). AutomationBench-AA measures agentic workflows across 657 tasks spanning 39 SaaS applications including Salesforce, Jira, Gmail, Zendesk and Google Sheets, scoring whether a model completes the objective without tripping a guardrail. DeepSeek V4.1 Flash gains 15 percentage points over DeepSeek V4 Flash 0731 (54%), sitting 12 points above DeepSeek V4 Pro 0813 (57%) and 7 points above GLM-5.3 (62%).

➤ DeepSeek V4.1 Flash is one of the most verbose models we’ve measured at 89k Tokens per Intelligence Index Task. That is 25% more than Z AI's flagship GLM-5.3 (71k), 29% more than GLM-5.3-Flash (69k), and 62% more than DeepSeek V4 Pro 0813 (55k). It uses more output tokens than even the frontier models, including Fable 5.1 (78k) and Claude Opus 5 (73k).

➤ Despite such verbosity, DeepSeek V4.1 Flash still costs just $0.27 per Intelligence Index task. This is primarily driven by its low pricing at $0.30 per 1M input tokens and $1.20 per 1M output tokens, with cached input tokens priced at just $0.006 per 1M tokens, a 98% discount. That is ~7x below both GLM-5.3 ( $2.01) and Kimi K3 ($2.00), and ~2.5x below DeepSeek V4 Pro 0813 ($0.67).

Additional model details:

➤ Context window: 1M tokens

➤ Pricing: $0.30 / $1.20 per 1M input/output tokens on DeepSeek's first-party API, with a 98% cache hit discount ($0.006 per 1M cached input tokens)

➤ Input modalities: text and image

➤ Size: This is a 763B total parameters, 8B active parameters input and 16B active parameters output

➤ License: MIT

➤ Providers: DeepSeek first-party API

[Token Harbor @TokenHarborAI]

Try DeepSeek V4.1 Flash ⚡

Model: `deepseek-v4.1-flash:free`
Base URL: `https://t.co/2tbXUhrK9v`

Free Models:
https://t.co/rvCnI06vuS

Passes + 2× Boost:
https://t.co/AcWPdeDH3G

[Token Harbor @TokenHarborAI]

DeepSeek V4.1 Flash is free on Token Harbor—but how does the limited 2× Model Boost work?

The quick answers:

• Free Tier access remains available
• Agent Pass: $0.99 first month → up to $20 usage
• Office Pass: $9.99/month → up to $70 usage
• The Model Boost ends September 17

After the boost ends, included usage returns to the regular Pass rate: $10 for Agent and $35 for Office.

[Token Harbor @TokenHarborAI]

@feldbarcelospro Hey Felipe, we’re doing great — thanks so much! That’d be awesome. Really glad Token Harbor feels like a good fit for Fractal!

[Token Harbor @TokenHarborAI]

⚡ $0.99 gets you $10 of AI usage.

Use GPT, DeepSeek, GLM, Qwen, MiMo and more.
Get even more usage with Model Boost.

Plus DeepSeek V4 Flash + MiMo V2.5 free.

[Token Harbor @TokenHarborAI]

@mengxu82835 2x boost lasts for one week only. the standard rate is $35

[Token Harbor @TokenHarborAI]

One week only: DeepSeek V4.1 Flash goes further on Token Harbor. ⚡

Agent Pass — $0.99 for your first month
Get up to $20 of DeepSeek V4.1 Flash usage.
Office Pass — $9.99/month
Get up to $70 of DeepSeek V4.1 Flash usage.

Free Tier is still available.

[Token Harbor @TokenHarborAI]

Already using DeepSeek V4 Flash?

Run the same prompt through deepseek-v4.1-flash:free and see which one wins 👀

[Token Harbor @TokenHarborAI]

Your first DeepSeek V4 Flash request in under 2 minutes ⚡

Create a free account
Copy your API key
Set base_url to https://t.co/2tbXUhrK9v
Use deepseek-v4-flash:free

No card required. Quick Start → https://t.co/dUtaZld1Ju

[Token Harbor @TokenHarborAI]

@sandra27tc4 yes, and you can expect more! we are planning a usage boost.

[Token Harbor @TokenHarborAI]

DeepSeek V4.1 Flash Beta is now free to try on Token Harbor. ⚡

New architecture.
Native multimodal.
Fast Flash-tier access.

No card required. Pick the deepseek-v4.1-flash:free and start building!

[Token Harbor @TokenHarborAI]

@martijnbry short outage , already fixed .

[Token Harbor @TokenHarborAI]

@deepseek_ai DeepSeek V4.1 Flash Beta scoring this high is kinda awkward for models that cost 10x more 😭

[Token Harbor @TokenHarborAI]

DeepSeek V4.1 Flash is officially released and free on Token Harbor. Here’s the 60-second setup ⚡

Base URL
https://t.co/2tbXUhrK9v
Model
deepseek-v4.1-flash:free

Works with the OpenAI SDK—change the base URL and model, then send your first request.

[DeepSeek @deepseek_ai]

🚀 Introducing DeepSeek-V4.1-Flash: smarter, faster, more efficient.

🔹 Introducing the smallest model in our new architecture family, with native visual understanding.
🔹 Designed for greater capability, faster inference, higher throughput, and scaling to larger models.

1/6

[Token Harbor @TokenHarborAI]

Check out our free model line:https://t.co/g72irFbHTg

Start building :https://t.co/otuJxi2KSp

[Token Harbor @TokenHarborAI]

Already using DeepSeek V4 Flash?
Run the same prompt through V4.1 Flash Beta and compare:
• output quality
• speed
• tool use
• multimodal handling

Both are free on Token Harbor:
deepseek-v4-flash:free
deepseek-v4.1-flash:free

[Token Harbor @TokenHarborAI]

@VonRealbuto2403 Yep, it’s legit. Feel free to try it out!