Recursive Flint 4.7x faster than Nemotron 3.5 🎃 We gave Flint, Ornith and Nemotron the same prompt in Atomic Local AI Agent’s orchestration mode, where Opus plans and the local models get the task done Try Atomic Fusion ->

MiniMax M3.1 Flash got 1.5x faster in Atomic Agent! We ran a Halloween prompt that builds one continuous 3D scene. In Atomic Fusion a local Qwen 3.8 27B wrote the plan, MiniMax did the coding and it was done in 8 minutes! Try Atomic Agent for FREE ->

Meet Atomic Agent Desktop! ✦ Local AI-first agent for Mac, Windows & Linux ✦ 4x bigger context window on local models with TurboQuant ✦ Atomic Fusion orchestrates cloud and local models to reduce costs Download Free & Private Agent →

Atomic Fusion cut GPT-6.1 Sol’s API bill by 77% 💥 We tested cloud-only, local-only and orchestrated builds. Cloud GPT-6.1 made a plan and orchestrated local Qwen 3.8 27B, making it cheaper with minimal quality loss Try Atomic Fusion ->

Atomic Agent connects 1500+ apps powered by @composio ✦ Set up with one API key ✦ Connect Gmail, Slack, Notion, GitHub and more ✦ Send emails, create tickets and manage files Available on Mac, Windows and Linux:

Atomic Agent v0.5.4 is out! ✦ Hugging Face and Ollama models ✦ Coding mode: plan, auto, or ask-first ✦ Context window: set the size of memory ✦ Fully redesigned TUI: scroll, click, select, copy ✦ Import data from Hermes, OpenClaw or ClaudeCode Built for local models:

Atomic Agent made GLM 5.3 better and it costs only 77 cents more! We ran three prompts: ✦ Teddy bear falling apart mid-run ✦ Frog popping ✦ Snowman melting in a sauna Outputs: ✦ GLM 5.3, API only: 543,290 tokens, $2.39 ✦ GLM 5.3 + Atomic Agent: 1,091,380 tokens, $3.16 An API call ends the moment the model stops typing. Atomic Agent keeps going: it runs the output, reads what came back, and hands the model its own failure to fix. The model picks the actions, the runtime owns the loop, the state and the stop conditions. The parts that matter here: ✦ one JSON array of tool calls per step, up to 15, reads running in parallel instead of one round trip per action ✦ results compressed before they go back into context, so the window holds the work instead of the transcript ✦ repeated actions get a warning at 3 and a veto at 5, so a stuck model stops instead of burning your budget in a circle

Qwen 3.8 27B on Atomic Agent was 2x faster than on Hermes and Prime agents! Outputs: Atomic: 5,535 lines, 19 modules, 2h35min Hermes: 10,424 lines, 109 modules, 4h15min Prime: 12,226 lines, 129 modules, 4h42min We gave three agents the same tasks, rebuild these games: ✦ Hole Io ✦ Tower Bloxx ✦ Subway Surfers Results: ✦ Atomic Agent gave us a real city going down the hole, skyscrapers tipping in, cars on the crossings, shops eaten one by one. In the tower every block swings in on its cable and lands where you drop it. Its runner is the cleanest of the three, though the world behind the boy stands still while he runs. ✦ Hermes dropped its hole onto a bare brown field with a few crates, and the hole grew while nothing ever fell in. Its tower buildings hang in the air with no cable at all, the logs reporting 12/12 tests passing on a game it had just broken. Its runner is an empty corridor with almost nothing in it. ✦ Prime built the prettiest town of the three, and five seconds in the whole thing tore off the ground into a tornado spinning around the hole, cars and houses circling forever and never falling in. Its tower lags on every drop, colours flickering on the buildings, and the lag is what brings the whole thing down by the fifth floor. In its runner the coins move at exactly the boy's speed, so the gap never closes and you can never pick one up. Run Qwen 3.8 27B in Atomic Agent!

Qwen 3.8 27B on Atomic Agent was 2x faster than on Hermes and Prime agents! Outputs: Atomic: 5,535 lines, 19 modules, 2h35min Hermes: 10,424 lines, 109 modules, 4h15min Prime: 12,226 lines, 129 modules, 4h42min We gave three agents the same tasks, rebuild these games: ✦ Hole Io ✦ Tower Bloxx ✦ Subway Surfers Results: ✦ Atomic Agent gave us a real city going down the hole, skyscrapers tipping in, cars on the crossings, shops eaten one by one. In the tower every block swings in on its cable and lands where you drop it. Its runner is the cleanest of the three, though the world behind the boy stands still while he runs. ✦ Hermes dropped its hole onto a bare brown field with a few crates, and the hole grew while nothing ever fell in. Its tower buildings hang in the air with no cable at all, the logs reporting 12/12 tests passing on a game it had just broken. Its runner is an empty corridor with almost nothing in it. ✦ Prime built the prettiest town of the three, and five seconds in the whole thing tore off the ground into a tornado spinning around the hole, cars and houses circling forever and never falling in. Its tower lags on every drop, colours flickering on the buildings, and the lag is what brings the whole thing down by the fifth floor. In its runner the coins move at exactly the boy's speed, so the gap never closes and you can never pick one up. Run Qwen 3.8 27B in Atomic Agent!

Open weight Kimi K3 performs at Claude Fable 5 level on 3D arcade games! We gave both models the same prompts: ✦ Pac-Man with four hunting ghosts ✦ Nokia Snake remade in 3D ✦ Retro space pinball table with real physics Outputs (thinking included): ✦ Kimi K3 (local, 8x B300): ~237K tokens, $0 ✦ Claude Fable 5: 92K tokens, $4.50 Every one of Kimi's tokens was free on our GPUs. Even via Moonshot's API the same Kimi run would be $3.60, still under Fable. Kimi held its quality against the strongest cloud model. Its snake is the liveliest scene of all six: soft shadows, floating dust, a flicking tongue, and a bot that checks it can still reach its own tail before every move. Fable's snake plays clean but looks flat next to it. Pac-Man is the closest call of the three. Kimi builds a fresh labyrinth on every restart and draws a live minimap in the corner while the red ghost tracks the player turn by turn through the corridors, and it even caught our bot on camera. Fable's maze answers with pearl pellets, +10 popups bursting over the floor and a wide-eyed cartoon ghost patrol. Fable won the pinball table: painted playfield art, a chrome ball that mirrors the neon lights, even little synth sound effects.

Open weight Kimi K3 performs at Claude Fable 5 level on 3D arcade games! We gave both models the same prompts: ✦ Pac-Man with four hunting ghosts ✦ Nokia Snake remade in 3D ✦ Retro space pinball table with real physics Outputs (thinking included): ✦ Kimi K3 (local, 8x B300): ~237K tokens, $0 ✦ Claude Fable 5: 92K tokens, $4.50 Every one of Kimi's tokens was free on our GPUs. Even via Moonshot's API the same Kimi run would be $3.60, still under Fable. Kimi held its quality against the strongest cloud model. Its snake is the liveliest scene of all six: soft shadows, floating dust, a flicking tongue, and a bot that checks it can still reach its own tail before every move. Fable's snake plays clean but looks flat next to it. Pac-Man is the closest call of the three. Kimi builds a fresh labyrinth on every restart and draws a live minimap in the corner while the red ghost tracks the player turn by turn through the corridors, and it even caught our bot on camera. Fable's maze answers with pearl pellets, +10 popups bursting over the floor and a wide-eyed cartoon ghost patrol. Fable won the pinball table: painted playfield art, a chrome ball that mirrors the neon lights, even little synth sound effects.

Atomic Agent beat Hermes on GAIA: 69.8% vs 58.5%, and it was 1.6x faster! We ran both agents through the full GAIA Level 1 benchmark, 53 real-world tasks, same 4-bit qwen-3.6-35b on the same Apple M4 Max. Results: ✦ Atomic Agent: 37 of 53 solved, done in 3h 12m ✦ Hermes Agent: 31 of 53 solved, took 5h 10m Atomic solved 6 more tasks and finished nearly 2 hours sooner. Hermes ran into the 900s timeout on 7 tasks; Atomic on just 2. Hermes burned 71% of its total time on tasks it still failed, Atomic, 48%. Where it showed: ✦ Audre Lorde poem, which stanza is indented: Atomic pushed through a dead source, switched tools, and answered in 7.6 min. Hermes ran the full clock and returned a blank. ✦ Vietnamese specimens, which city they ended up in: Atomic pulled it from the first source and normalized the answer in 33s. Hermes spent 7.3 min and never answered. ✦ The dinosaur featured-article nominator: Atomic walked the Wikipedia chain to "FunkMonk" in 57s. Hermes guessed a wrong name after 11 min. Atomic keeps a byte-stable prompt prefix, so llama-server reuses the KV-cache instead of re-encoding the whole context every turn, and it emits one JSON array of tool calls per inference, then compresses results back instead of pasting them in full, so the context never balloons and a small model stays sharp deep into a task. On top of that a no-progress guard vetoes repeated identical tool calls (warn at 3, hard veto at 5) and forces a reply, so Atomic never sinks 15 minutes into re-scanning one page the way Hermes did. Both agents missed some of the same questions, and on a few Hermes got there and Atomic did not, usually format slips where Atomic computed the right number but printed the working instead of the bare value. But on identical hardware and identical weights, the runtime that reuses its cache and refuses to spin came out ahead on accuracy and speed. Getting this from the runtime alone is wild. Run the same 53 GAIA tasks on Atomic Agent!

Atomic Agent beat Hermes on GAIA: 69.8% vs 58.5%, and it was 1.6x faster! We ran both agents through the full GAIA Level 1 benchmark, 53 real-world tasks, same 4-bit qwen-3.6-35b on the same Apple M4 Max. Results: ✦ Atomic Agent: 37 of 53 solved, done in 3h 12m ✦ Hermes Agent: 31 of 53 solved, took 5h 10m Atomic solved 6 more tasks and finished nearly 2 hours sooner. Hermes ran into the 900s timeout on 7 tasks; Atomic on just 2. Hermes burned 71% of its total time on tasks it still failed, Atomic, 48%. Where it showed: ✦ Audre Lorde poem, which stanza is indented: Atomic pushed through a dead source, switched tools, and answered in 7.6 min. Hermes ran the full clock and returned a blank. ✦ Vietnamese specimens, which city they ended up in: Atomic pulled it from the first source and normalized the answer in 33s. Hermes spent 7.3 min and never answered. ✦ The dinosaur featured-article nominator: Atomic walked the Wikipedia chain to "FunkMonk" in 57s. Hermes guessed a wrong name after 11 min. Atomic keeps a byte-stable prompt prefix, so llama-server reuses the KV-cache instead of re-encoding the whole context every turn, and it emits one JSON array of tool calls per inference, then compresses results back instead of pasting them in full, so the context never balloons and a small model stays sharp deep into a task. On top of that a no-progress guard vetoes repeated identical tool calls (warn at 3, hard veto at 5) and forces a reply, so Atomic never sinks 15 minutes into re-scanning one page the way Hermes did. Both agents missed some of the same questions, and on a few Hermes got there and Atomic did not, usually format slips where Atomic computed the right number but printed the working instead of the bare value. But on identical hardware and identical weights, the runtime that reuses its cache and refuses to spin came out ahead on accuracy and speed. Getting this from the runtime alone is wild. Run the same 53 GAIA tasks on Atomic Agent!