DGrid AI

💡Gemini 4 Carbon just dropped on DGrid AI.

@Google’s strongest model — for the problems that push past Argon.
🔹 Built for long-horizon and multi-step reasoning
🔹 Checks its code and the screen as it builds, from complete apps to 3D worlds
🔹 Flags missing or flawed data instead of guessing
🔹 Same intro pricing as Argon: $2 / $10 per 1M tokens

Try it now:
https://x.com/dgrid_ai/status/2120721719292039475

🎉 Excited to team up with @Didotxyz_!

Built on a crypto-native model, DI's AI + crypto intelligence ecosystem connects market data, event research, quant products and interoperable agents. ⚡️

DGrid'll support DI’s Agent Network with an efficient, low-cost #AI API for model access and verifiable inference.

Full post:
https://x.com/dgrid_ai/status/2120063922711507374

⚡️Claude Fable 5.2 from @claudeai is now on DGrid AI.

Built for the most ambitious technical work. A step above Opus 5.5 on nearly every capability Anthropic tested, and up to 30% faster than Fable 5.1.
🔸 Improvements in coding, reasoning, and vision
🔸 Clearer communication than previous Fable models 🔸 On FrontierCode 2.0, High effort matches Opus 5.5 Max at roughly half the cost per task

Try it now:
https://x.com/dgrid_ai/status/2119992036460019855

🚨 AI is booming, but trust remains a core issue — users still can’t easily verify they’re getting the model and quality they were promised.

🫡 Anthropic’s FlashProve is an important step toward verifiable AI, cryptographically proving that a response came from the claimed model.

In multi-model, multi-provider environments, we take a complementary path with PoQ(Proof of Quality): multi-dimensional quality assessment + on-chain verifiable proofs to continuously check that the service matches what was promised.👀 Different technical paths, same destination: verified AI.

📑 More on our PoQ research:
https://x.com/dgrid_ai/status/2118976755650253137

✨New image models from @OpenAI just landed on DGrid AI.

GPT-Image-2.6 Flare and Sunburst — 50% faster generation overall.
🔸 Flare generates images in as little as six seconds
🔸 Sunburst delivers higher quality and better consistency across edits

Image markup and SVG exports also included.

Try them now:
https://x.com/dgrid_ai/status/2118606012991336451

✨New image models from @OpenAI just landed on DGrid AI.

GPT-Image-2.6 Flare and Sunburst — 50% faster generation overall.
🔸 Flare generates images in as little as six seconds
🔸 Sunburst delivers higher quality and better consistency across edits

Image markup and SVG exports also included.

Try them now:
https://x.com/dgrid_ai/status/2118606012991336451

💡 Our sixth paper is out!

FIDES-Live: Evaluating Behavioral Concordance in Sequential Trading Agents A

n AI trading agent can submit a valid order, have it rejected, and still report that the trade was completed. A correct account balance alone would not reveal this discrepancy.

In this paper, we examine whether an agent’s decisions, proposed orders, execution outcomes, and subsequent reports agree with one another. We introduce FIDES-Live to audit this behavioral concordance through the records produced during each trading round.

We audited 10 requested inference routes across 7,609 account-rounds in cryptocurrency and an equity pilot, using real market observations and local paper execution.

Key findings from the cryptocurrency traces:

🔹 Decision validity reached 97.35%, while structured reporting accuracy was 90.81%, among scorable responses.
🔹 Commitment adherence was 94.36% overall, but only 62.63% for constraints declared with nonempty order proposals.
🔹 Only 25.13% of valid decisions declared a nontrivial commitment.

Coverage and activity matter when interpreting adherence. Across both environments, every retained round’s cash and positions reconciled with recorded fills. Yet all 10 routes had negative marked returns in the settled cryptocurrency session.

✅ Valid decisions do not guarantee accurate reports. ✅ Ledger reconciliation does not establish profitable trading.

Trading agents need to be evaluated across decisions, commitments, execution, and reporting, with financial outcomes measured separately.

Full Paper:
https://x.com/dgrid_ai/status/2118595864004563368

🚨DGrid Model Marketplace Now Processes 60B Tokens and 2M+ API Calls Every Day

A shared achievement for everyone building with us.🥳

DGrid Model Marketplace now handles production-scale AI workloads every day. Behind these numbers are providers making their models available, developers building them into products, and #AI Agents running real tasks around the clock.

🔸2M+ API calls per day
🔸60B tokens processed per day
🔸200+ models available
🔸100+ providers on the marketplace

#DGrid Gateway routes real demand across 100+ providers, up from 30+ earlier this year. Agent workflows and long-context applications account for most of the marketplace’s daily token usage.

To every provider and developer who helped us get here: thank you. Every model listed and application shipped helps this marketplace grow.🧡💛

Next, we’ll improve model discovery, integration, and inference performance—helping providers reach more users and developers move faster from first API call to production.

Built on real usage. And we’re just getting started.

Full post:
https://x.com/dgrid_ai/status/2117900903588196859

Claude Code × DGrid #AI Gateway ⚡️

From your #DGrid API key to your first successful response—every step, documented. Our integration guide covers CLI and desktop setup, with configuration examples and troubleshooting along the way.

👇Connect your workflow.
https://x.com/dgrid_ai/status/2117531743557447973

🚨 Kimi K3.1 Pro and K3.1 Pro Highspeed from@Kimi_Moonshot are now on DGrid AI.

Moonshot’s most capable 3T-class open-weight model — 1M context, multimodal, and matching the strongest proprietary models on coding, agentic, and knowledge work.

✨ Full-project coding: games in Unreal, Unity & Godot ✨ Strong at parametric 3D modeling in Blender & CAD ✨ More reliable finance work; hallucination down 65% ✨ Highspeed version runs 2–3× faster

Try them now:
https://x.com/dgrid_ai/status/2113896961707573707

🔥 DeepSeek-V4.1-Flash-1017 from @deepseek_ai
just landed on DGrid AI.

Stronger benchmarks, faster task completion, and shorter outputs than the previous version — a Pareto improvement in performance and cost.

🔸 Better end-to-end project work, from game creation to 3D modeling
🔸 Optimized for DeepSeek Harness and coding agents 🔸 Open weights available

Try it now:
https://x.com/dgrid_ai/status/2113163439908561248

🧡 Excited to join The Blockchain 100 as an Ally Partner!

Looking forward to celebrating the creators driving the space forward with @binance⚡️

Full post:
https://x.com/dgrid_ai/status/2112375785399173443

🥳 Introducing DGrid Auto Deposit.

Set your trigger balance, target balance, approved limit, and expiry date. When your balance runs low, #DGrid tops it up from your authorized wallet, helping long-running #AI tasks, services, and workflows stay funded and running.

👇Tutorial:
https://x.com/dgrid_ai/status/2111002540011900979

🚨 $DGAI is live in the #Binance Alpha Trading Competition.

Let’s go. ⚡️

More info:
https://x.com/dgrid_ai/status/2110347523986805058

https://x.com/BinanceWallet/status/2110325836549414948

🔥 Gemini 4 Argon is now on DGrid AI.

@Google’s new frontier model, built to take complex ideas from start to finish.

✅ Breaks tasks down step-by-step and helps produce real outputs: shareable web apps, ready-to-print 3D files, slides, and reports
✅ Major improvements in coding
✅ Strong on complex, multi-step work across software engineering and knowledge work

Try it now:
https://x.com/dgrid_ai/status/2110215614438777198

🙌 Exactly. Routing to the right model for the task is table stakes now.

At @dgrid_ai we built a Unified Gateway & Smart Routing — a millisecond-level engine that automatically routes prompt intents to the optimal global compute resources based on task type, budget, and latency.

One API. The best fit, every time.

Full post:
https://x.com/dgrid_ai/status/2109913914058166534

👀 Think you can make the Top 50?

The DGrid AI Arena Weekly Leaderboard is waiting. Every week, the Top 50 players share 50,000 points, with bigger rewards for higher ranks.

👇 Enter the Arena, compete for your spot, and see how high you can climb:
https://x.com/dgrid_ai/status/2109541019519996289

🚀@Google's Nano Banana 2.1 is live on DGrid AI.

Their newest image generation and editing model. Stronger on visual design, mask-based editing, and keeping subjects consistent, for more natural results.

Try it now:
https://x.com/dgrid_ai/status/2108803507663045067

🔥 Claude Haiku 5.5 from @claudeai is now on DGrid AI.

@AnthropicAI 's cheapest, fastest, and most capable small model yet. On average, it costs around 75% less to run than Haiku 4.5.
⚡️ A step up over Haiku 4.5 on coding, computer use, and knowledge work
⚡️ Built for high-volume work: summaries, classification, live support, and browser use
⚡️ Pairs well with Opus 5.5 and Sonnet 5.5 as a coding subagent

Try it now:
https://x.com/dgrid_ai/status/2108041511648768367

🌟We’re live at @token2049 Singapore!

Come meet us at Booth MB4-80. See you there!⚡️

Full post:
https://x.com/dgrid_ai/status/2107662930238194127