Get the Agent OS + join 4,000+ founders inside the AI Profit Boardroom → Join AIPB ($59/mo)
Moonshot AI

Kimi K3

Moonshot's 2.8T flagship — 1M context, tuned for long-horizon agent work.

Context1,048,576 tokens — a full codebase in working memory
Pricing$3 / M in
Tasks tested50
Avg score7.89/10 average
Medals🥇8 🥈4 🥉7
Release2026-07-16
Official sitekimi.com ↗
Official vendor source
Kimi K3 is built by Moonshot AI — see the vendor's own product page, pricing, and docs at kimi.com.
Visit kimi.com →

Reference benchmarks for Kimi K3

These are external benchmarks I pulled from the source comparison guides on agentos.guide — SWE-bench Verified, DRACO, Kilo plan rubric, build-time measurements, vendor-reported coding scores. They are not goldiebench medal scores (those come only from same-prompt one-shot creative coding tasks in the matrix). I surface them here so the spec sheet for Kimi K3 is honest about what's measured.

Launch-day community benchmarks
~Fable/Sol tier; Terminal Bench the standout

What is Kimi K3?

Kimi K3 is the Moonshot AI frontier model with a 1,048,576 tokens — a full codebase in working memory context window, released 2026-07-16. Tagline: Moonshot's 2.8T flagship — 1M context, tuned for long-horizon agent work.. Official source: kimi.com.

Pricing detail. Launched July 16, 2026. 2.8T-param MoE (Moonshot's quickstart corrected the circulating 2.5T estimate). $3/M input on OpenRouter at launch; included at no extra cost in the Kimi coding plan (`k3` on the coding endpoint).

How I use it inside the Agent OS. Wired into the Agent OS as the `kimi-k3` Hermes profile and a K3 speed-toggle in the Kimi Code tab — used for long unattended agent runs where a slow-but-right model beats a fast-but-forgetful one.

What I built with Kimi K3

Every model on Goldie Bench gets the same fixed prompt set — one shot, single HTML file out — and I score the result 0–10 inside the Agent Operating System. Here's what Kimi K3 shipped on the bench: 50 one-shot demos across 1,048,576 tokens — a full codebase in working memory of context. Of those, 50 are scored against the field with my honest 0–10 from the source guides at agentos.guide.

Strengths

  • Launch-day benchmarks put it around the Fable/Sol tier, with Terminal Bench (agentic terminal-driving) the standout
  • 1M-token context verified on this bench's needle test: exact recall from 162k tokens of noise in 18s
  • One-shot builds run long but land complete — its first bench game (13.4 min of thinking, 30,880 tokens) playtested with zero JS errors
  • Included in the Kimi coding plan — frontier tier without a new bill

Trade-offs

  • Slow on hard tasks — early testers report up to ~35 minutes at max reasoning; this bench saw 13+ minute single builds
  • Launch-day rate limits on OpenRouter (429s) — the coding-plan endpoint was the reliable route
  • Self-reports as K2.7 if you ask it — verify the served model via the API response, not the model's word

Best for

  • long-horizon agent runs
  • whole-repo context work
  • terminal-driving agents

Every demo by Kimi K3

50 live demos, sorted by category. Click any tile to play the actual one-shot result. Verdicts and 0–10 scores are pulled from the source guides where I posted them publicly.

Kimi K3 one-shot build of Crypt — GoldieBench AI benchmark screenshot▶ LIVE
Crypt
Game
Renders a clean, atmospheric brick corridor with polished title typography, a working minimap, and solid HUD framing, but the scene reads cool-blue and under-lit rather than warmly torch-lit — no visible flames or orange glow undercut the core brief, keeping it strong-but-generic rather than a task winner.
Kimi K3 one-shot build of Dragonrealm — GoldieBench AI benchmark screenshot▶ LIVE
Dragonrealm 🥉
Game
Strong atmospheric render nails the Skyrim-style frozen night — moon, aurora, snowfall, drawn sword in-hand, glowing braziers, distant dragon silhouette on compass, and a full HUD with health/stamina bars. Polished lighting and moody vignette push it to the top of the field; minor nitpick is somewhat sparse mid-field geometry, but on-brief execution is excellent.
Kimi K3 one-shot build of Doom — GoldieBench AI benchmark screenshot▶ LIVE
Doom
Game
Strong, polished raycaster with detailed procedural lava/metal wall textures, minimap, HUD, gun sprite and a distant medkit pickup visible — very shippable DOOM atmosphere. Weak point: no monster is visible in the rendered frame and HP already at 91, so the core 'monsters chase you' brief isn't confirmed on screen, keeping it just below the top tier.
Kimi K3 one-shot build of Racing — GoldieBench AI benchmark screenshot▶ LIVE
Racing
Game
Strong, polished night racer with curving road, kerbs, dashed lanes, mountains, trees, striped barrier obstacles, and a clean HUD (speed/score/best) plus a CRASH state — very shippable. Slight ding: the car itself appears to render only as wheels + shadow blob (body missing from view), which hurts the third-person presentation just short of task-winner.
Kimi K3 one-shot build of Skyrim — GoldieBench AI benchmark screenshot▶ LIVE
Skyrim
Game
Renders a genuinely atmospheric open world with layered fog, warm low sun, distant mountains, a village of houses, scattered pines, plus a polished serif HUD with scrolling compass and stamina bar — clearly on-brief and shippable. The oversized green glow sprites (fireflies/particles) look blown-out and slightly buggy in the foreground, holding it just shy of top-tier.
Kimi K3 one-shot build of Twilightvale — GoldieBench AI benchmark screenshot▶ LIVE
Twilightvale
Game
Renders a coherent twilight low-poly world with cohesive purple/gold theming, full HUD (minimap, XP/HP/stamina bars, quest chip, NPC dialogue), trees, houses and water — clearly on-brief and shippable. However the camera clips badly into an oversized foreground terrain spike that dominates the frame, and the flat untextured cube houses look crude, keeping it below the polished top tier.
Kimi K3 one-shot build of Voxelcraft — GoldieBench AI benchmark screenshot▶ LIVE
Voxelcraft 🥉
Game
Strong: gorgeous procedural terrain with textured grass/dirt/stone/water/trees, real shadows, day/night clock HUD, crosshair and control hints all rendering cleanly and on-brief. Weak: the hotbar area at bottom looks faded/half-rendered in the shot, but the world itself is clearly a top-tier voxel sandbox that matches the field's best.
Kimi K3 one-shot build of Gtafoot — GoldieBench AI benchmark screenshot▶ LIVE
Gtafoot 🥉
Game
Strong, atmospheric neon night city with a well-composed third-person character, clean HUD (wanted stars, health, ammo, minimap, money) and full weapon/cover/pedestrian systems in source; visually polished and clearly on-brief, but the emissive-window city and blocky avatar feel a notch below the field's best rather than a clear task winner.
Kimi K3 one-shot build of Gtadrive — GoldieBench AI benchmark screenshot▶ LIVE
Gtadrive 🥉
Game
Strong atmospheric render — cohesive dusk/neon skyline, lit facade textures, on-foot character with steal prompt, circular minimap with compass, and clean HUD all deliver the GTA sandbox brief. Loses a touch for a slightly empty foreground and unverified cop-chase/wanted mechanics, but visually polished and clearly on-brief.
Kimi K3 one-shot build of Aipbpromo — GoldieBench AI benchmark screenshot▶ LIVE
Aipbpromo
Page
The 3D wireframe starfield background renders beautifully and the HUD (progress bar, scene tags, dots, skip) is polished, but the screenshot captured a PAUSED state showing no actual scene content—no logo, stats, or CTA visible—leaving the core cinematic brief unproven in the render.
Kimi K3 one-shot build of Parachute — GoldieBench AI benchmark screenshot▶ LIVE
Parachute
Game
Renders cleanly with a polished 3D scene—gradient sky, mountains, clouds, altitude bar, sink/wind stats, and a pulsing PULL button—clearly on-brief for freefall/steer/pull/land. Strong HUD and controls, but the jungle canopy reads as flat white/gray squares rather than dense green trees and the target clearing isn't visible from this angle, keeping it just below the top tier.
Kimi K3 one-shot build of Flightsim — GoldieBench AI benchmark screenshot▶ LIVE
Flightsim
Game
Strong, polished render with a clean 3D plane on a proper runway, full HUD (compass ribbon, artificial horizon, IAS/ALT gauges, throttle slider) and nice terrain/hangar/windsock props; minor blemish is the 'VS NaN' readout indicating an uninitialized vertical-speed calc, which keeps it just below the top tier.
Kimi K3 one-shot build of Arcade — GoldieBench AI benchmark screenshot▶ LIVE
Arcade
Game
Gorgeous neon breakout that rendered cleanly with rainbow bricks, glowing paddle, ball trail, starfield and CRT scanlines, plus a full feature set (powerups, multiball, levels, audio, auto-demo). Very close to the field's best; loses a hair for being a well-trodden brief rather than a novel take.
Kimi K3 one-shot build of Dogfight — GoldieBench AI benchmark screenshot▶ LIVE
Dogfight
Game
HUD, title screen, and crosshair are polished and on-brief, but the 3D scene failed to render — the background is pure black despite code defining a sky dome, ground, clouds, and planes, so no actual dogfight/flight is visible.
Kimi K3 one-shot build of Dragonflight — GoldieBench AI benchmark screenshot▶ LIVE
Dragonflight
Game
Strong 3D scene renders cleanly: a modeled dragon flying through glowing neon rings over a synthwave terrain with full HUD (score, rings, speed, combo, fury meter) all present and on-brief. Slightly loses points—fire-breath isn't visible in the shot and the dragon model reads a bit plain/spherical up close—but it's polished, cohesive, and clearly matches the brief.
Kimi K3 one-shot build of Game — GoldieBench AI benchmark screenshot▶ LIVE
Game
Game
Strong, polished 3D synthwave endless runner with Three.js—glowing player orb, neon rails, moving pylons, starfield sun, and a clean HUD (score/best/lives/speed) all render beautifully and on-brief. Deducts slightly for being a familiar lane-runner concept, but the execution and neon aesthetic push it near the top of the field.
Kimi K3 one-shot build of Neonblaster — GoldieBench AI benchmark screenshot▶ LIVE
Neonblaster
Game
Clean neon aesthetic with glowing title, HP diamonds, bomb button, star grid backdrop and a polished 'WAVE 1' banner render; source shows genuinely deep systems — 5 enemy types, bosses with named patterns, power-ups, procedural synth music with delay/compressor, screen-shake — but the screenshot itself is fairly sparse mid-announce so it reads slightly less juicy than the top of field.
Kimi K3 one-shot build of Neoncity — GoldieBench AI benchmark screenshot▶ LIVE
Neoncity 🥈
Game
Gorgeous on-brief cyberpunk drive: glowing multicolor building windows, cyan/magenta neon road edges, a synthwave sunset moon down the vanishing point, and clean HUD with speed/distance—strongly atmospheric and polished. Minor nitpicks are the somewhat sparse traffic and generic sign legibility, but the render clearly nails the vaporwave neon-city vibe at a task-winning level.
Kimi K3 one-shot build of Neonracer — GoldieBench AI benchmark screenshot▶ LIVE
Neonracer 🥈
Game
Gorgeous, fully-on-brief synthwave scene — retro sun, wireframe mountains, neon grid, dashed lanes and glowing vapor-trail particles behind the car all render crisply with polished HUD. Only minor nitpick is the car model reads a touch plain, but visually this matches/beats the field's best.
Kimi K3 one-shot build of Nordiccrypt — GoldieBench AI benchmark screenshot▶ LIVE
Nordiccrypt
Game
Strong: genuine first-person 3D dungeon with convincing torch glow, warm falloff lighting on stone brickwork, scattered rubble props, and a polished serif Nordic UI with rune-collection goal that fully nails the brief. Minor weakness: the corridor geometry is a bit boxy and props read as simple low-poly, but the atmosphere and rendering are clearly task-winning quality.
Kimi K3 one-shot build of Outrun — GoldieBench AI benchmark screenshot▶ LIVE
Outrun
Game
Gorgeous rendered scene — gradient sunset with scanlines, layered mountain ridges, glowing pyramids/palms/signs, and clean pseudo-3D road with dashed lanes and traffic all read perfectly on-brief. Only nit is the player car's flat pink slab looks slightly less refined than the atmospheric backdrop, but the overall polish clears the field's best.
Kimi K3 one-shot build of Pool — GoldieBench AI benchmark screenshot▶ LIVE
Pool
Game
Renders a polished 3D pool table with proper rack, numbered balls, cue stick, aim line, pockets, HUD and physics/audio scaffolding — clearly shippable and on-brief. Weak points: the floating lamp shade looks unlit/misplaced and the rack triangle is slightly loose, keeping it just below the top tier.
Kimi K3 one-shot build of Raycaster — GoldieBench AI benchmark screenshot▶ LIVE
Raycaster 🥇
Game
Strong render: multiple distinct textured wall types (brick, blue stone, glowing tech strips, mossy hedge), textured floor/ceiling, minimap, ceiling glow lights and vignette all working with clean UI and controls. Very polished and clearly on-brief; edges out the field with texture variety and emissive detail, though sprites/enemies aren't visible in this frame.
Kimi K3 one-shot build of Rpg — GoldieBench AI benchmark screenshot▶ LIVE
Rpg
Game
Strong, polished render: cohesive pixel-art sprites (knight, skeletons, bats), varied grass/flower/path tiles, trees and rocks, plus a clean HUD with HP/level/gold/DMG, minimap and slain counter — clearly on-brief with combat, inventory and leveling systems in source. Loses a touch versus the field's best on world visual variety and the fairly barren procedural layout, but it's genuinely shippable.
Kimi K3 one-shot build of Matrixrain — GoldieBench AI benchmark screenshot▶ LIVE
Matrixrain 🥈
Other
Renders beautifully with dense katakana streams, glowing heads, mid-trail glyphs, a clean flickering title and CRT scanline/vignette overlay; strong extras like 6 themes, speed control, pause overlay, and mouse-disturb interaction push it above the field's best.
Kimi K3 one-shot build of Mlx Speedtest — GoldieBench AI benchmark screenshot▶ LIVE
Mlx Speedtest
Other
Strong 3D-backed speedtest with polished neon gauge, live stats, and orbiting torus knot that renders cleanly and reads as on-brief; weakened by the busy wireframe knot obscuring the central gauge/readout and the sparkline area appearing as an empty blurred bar, reducing legibility.
Kimi K3 one-shot build of Neonsnake — GoldieBench AI benchmark screenshot▶ LIVE
Neonsnake 🥇
Other
Strong 3D take on the brief: crisp neon walls with glowing corner nodes, tilted perspective grid, glowing snake with eyes and a radiant food orb, plus polished HUD, autopilot AI, and pause overlay — visibly beats a flat 2D field. Minor nit is the slightly empty arena interior, but the render is clean, on-brand, and clearly a task winner.
Kimi K3 one-shot build of Landing — GoldieBench AI benchmark screenshot▶ LIVE
Landing
Page
Strong: the live 3D floating geometry over a particle-grid backdrop is genuinely striking and elevates a standard SaaS layout, with polished gradient type, marquee logo strip, and clean feature cards. Weak: the hero headline slightly overlaps the 3D torus, but the overall polish and completeness (nav, logos, features, stats, pricing, FAQ) make it a task-winning build.
Kimi K3 one-shot build of Webos — GoldieBench AI benchmark screenshot▶ LIVE
Webos
Page
Strong, highly polished render: crisp macOS-style traffic-light windows, blurred glass dock with running dots, desktop icons, animated 3D wireframe backdrop and starfield, plus a clean welcome/about card — clearly on-brief with Notes/Paint/Terminal. Source confirms real window management (drag/resize/min/max), autosave, and functional apps; only mild risk is unseen app depth, but presentation and completeness edge past the field's best.
Kimi K3 one-shot build of Blackhole — GoldieBench AI benchmark screenshot▶ LIVE
Blackhole 🥇
Sim
Gorgeous real geodesic raytracing with a properly lensed accretion disk wrapping over the top, convincing Doppler asymmetry, photon-ring glow, and a rich nebula starfield backdrop; polished typography and clean UI push it to the top of the field. Minor stat overlap at bottom-left ('photon sphe' clipped by hint pill) is the only blemish.
Kimi K3 one-shot build of Boids — GoldieBench AI benchmark screenshot▶ LIVE
Boids 🥇
Sim
Renders a polished 3D boids sim with genuine emergent flocking clusters, colorful instanced birds with wings, a glowing repel/attract orb, dusk gradient sky and a clean control panel — strong visual and simulation depth that matches the top of the field. Minor weakness: 44fps at 280 boids hints at some performance cost, but it's clearly a task winner.
Kimi K3 one-shot build of Cloth — GoldieBench AI benchmark screenshot▶ LIVE
Cloth
Sim
Strong render: the checkered cloth drapes convincingly over the sphere with realistic folds, soft shadows, and clean UI; loses a touch because the underlying pedestal/sphere object is fully hidden and the scene reads slightly flat rather than showing the object being draped.
Kimi K3 one-shot build of Fluid — GoldieBench AI benchmark screenshot▶ LIVE
Fluid 🥇
Sim
Gorgeous, textbook-quality WebGL fluid sim with rich swirling dye, added particle sparkle, and polished UI (gradient title, hint pill, control buttons) — the vorticity/pressure-solve pipeline and half-float fallback handling are all correct and shippable; only minor knock is the visual is close to the well-known Pavel Dobryakov reference.
Kimi K3 one-shot build of Fractal — GoldieBench AI benchmark screenshot▶ LIVE
Fractal
Sim
Crisp GPU-shader Mandelbrot renders beautifully with smooth continuous coloring, vivid Aurora palette, and a polished glassy HUD; full feature set (mode toggle, Julia spawn, palettes, zoom/pan, keyboard) makes it a task winner, only mild nit being the intense magenta background dominating the frame.
Kimi K3 one-shot build of Galaxy — GoldieBench AI benchmark screenshot▶ LIVE
Galaxy
Sim
The UI chrome (title, buttons, hint bar) renders beautifully with polished gradient typography, but the core galaxy particle system is completely absent — the canvas is empty black and '—PARTICLES' shows the count never populated, indicating the Three.js render failed or hung. A stylish shell around a non-functional centerpiece.
Kimi K3 one-shot build of Orbit — GoldieBench AI benchmark screenshot▶ LIVE
Orbit
Sim
The UI chrome (title, hint, control panel) renders beautifully, but the entire simulation canvas is empty black — no star, planets, asteroid belt, or trails appear, so the core N-body simulation failed to render. Strong styling can't compensate for a non-functioning main feature.
Kimi K3 one-shot build of Particleforge — GoldieBench AI benchmark screenshot▶ LIVE
Particleforge
Sim
Polished chrome — gradient title, presets, gravity slider, glowing core and 61fps at 14k particles all read well, and the trails/orbiter system is real. But the screenshot shows only a sparse thin ring of orbiters rather than a dense swirling forge of particles, so the core 'sculpt swirling systems' visual feels underwhelming and empty compared to the field's best.
Kimi K3 one-shot build of Pathtracer — GoldieBench AI benchmark screenshot▶ LIVE
Pathtracer 🥉
Sim
Genuine GLSL Monte Carlo path tracer rendering a proper Cornell box with correct colored-wall bleeding, metal/glass/diffuse spheres, soft shadows and a refracting glass sphere with visible caustics — clearly physically-based and progressively converging (samples counter live). Only weakness is the noisy, under-converged framebuffer and the header text overlapping a reflected sphere, but the physics and polish edge it past the field's best.
Kimi K3 one-shot build of Reactiondiff — GoldieBench AI benchmark screenshot▶ LIVE
Reactiondiff
Sim
Strong UI polish with gradient title, glassy HUD, presets and sliders, but the screenshot shows only isolated seed blobs rather than the emergent Turing patterns the sim should grow — it appears paused (Run button showing) so the reaction-diffusion character isn't visible, undermining the core brief.
Kimi K3 one-shot build of Solar — GoldieBench AI benchmark screenshot▶ LIVE
Solar
Sim
Strong UI chrome (glass panels, chips, info card, speed clock) and rich procedural textures, but the rendered view is visually muddy — an oversized washed-out sun bleeds a brownish haze across the whole scene, orbit rings barely read, and the focused 'Mercury' shows a giant gray moon-like body while ringed Saturn floats tiny top-left, breaking the sense of accurate scale/placement. Polished framework but the actual 3D render is confusing and off-brief for an accurate system.
Kimi K3 one-shot build of Wormhole — GoldieBench AI benchmark screenshot▶ LIVE
Wormhole 🥉
Sim
Gorgeous, on-brief render — concentric glowing gate rings receding into a bright vanishing point, colorful volumetric plasma glow and floating energy particles sell real tunnel depth, with a polished HUD/title. Only nit is the hard diagonal seam in the upper-right that slightly breaks the illusion, but overall this matches/beats the field's best.
Kimi K3 one-shot build of Aurora — GoldieBench AI benchmark screenshot▶ LIVE
Aurora 🥇
Visual
Gorgeous flowing volumetric aurora ribbons with convincing fbm noise, layered mountains, spruce silhouettes, moon, stars and a shooting star make a genuinely atmospheric scene; the elegant typography, palette switcher and vignette give it a shippable polish that edges past the field's best.
Kimi K3 one-shot build of Fireworks — GoldieBench AI benchmark screenshot▶ LIVE
Fireworks
Visual
Gorgeous rendered scene with a detailed city skyline, moon, stars, and a rich variety of firework shapes (rings, willows, flowers, hearts, chrysanthemums) all glowing convincingly; the polished HUD and multiple simultaneous bursts make it a task winner, with only slight visual noise from overlapping trails as a minor nit.
Kimi K3 one-shot build of Lavalamp — GoldieBench AI benchmark screenshot▶ LIVE
Lavalamp 🥉
Visual
Strong: a genuinely convincing lava lamp silhouette with glowing base, heated cap, warm wax gradient and a proper metaball shader plus polished typography and theme chips; minor weakness is the blob field looks a touch sparse/stratified in this frame, but the craft and shader detail clearly edge past the field's best.
Kimi K3 one-shot build of Matrix — GoldieBench AI benchmark screenshot▶ LIVE
Matrix 🥈
Visual
Screenshot shows crisp katakana glyph rain with bright leading heads fading trails, a glowing mouse-pulse orb, scanline/vignette CRT treatment and a strong glowing title — clearly on-brief and polished. Rich interactivity (6 themes, click bursts, mouse gusts, pause/fullscreen) and multi-layer depth push it to task-winner tier.
Kimi K3 one-shot build of Plasma — GoldieBench AI benchmark screenshot▶ LIVE
Plasma
Visual
Clean UI with gradient PLASMA title, numbered palette swatches, and clear hint text render nicely; however the actual plasma field looks like soft amber blobs rather than the rich, banded hypnotic plasma the brief calls for, making it visually flat compared to the field's best. Solid, shippable but not a task winner.
Kimi K3 one-shot build of Synthwave — GoldieBench AI benchmark screenshot▶ LIVE
Synthwave 🥇
Visual
Nails every synthwave trope beautifully — banded sunset, receding neon perspective grid, layered mountains, glowing gradient title, floating wireframe solids, starfield, scanlines/vignette, and generative WebAudio music. Highly polished and cohesive; a clear task winner.
Kimi K3 one-shot build of Terrain — GoldieBench AI benchmark screenshot▶ LIVE
Terrain
Visual
Strong scaffolding — nice sky gradient, water, clouds, sand beaches, instanced trees, and clean UI — but the terrain interior renders almost entirely black, indicating a broken lighting/normals or vertex-color issue that dominates the screenshot. The visible flaw undercuts an otherwise polished procedural island explorer.
Kimi K3 one-shot build of Voxel — GoldieBench AI benchmark screenshot▶ LIVE
Voxel
Visual
Strong render — clean voxel island with sandy shores, translucent water, lush grass terrain, trees, and scattered flowers, plus polished HUD, clouds, shadows, and interactive controls (orbit/zoom/mine/regen). Feels genuinely Minecraft-style and shippable; minor nit is slightly flat lighting variety but overall a task-topping build.
Kimi K3 one-shot build of Waves — GoldieBench AI benchmark screenshot▶ LIVE
Waves 🥇
Visual
Stunning render — layered Gerstner waves with convincing foam crests, a glowing sun with a rippling specular light path across the water, and a bobbing sailboat riding the swell adds narrative life. The teal-to-deep-blue palette, sun glow sprite, and clean typographic UI make it clearly a task-topping build.
every demo, in a grid · click any one to play

Compare Kimi K3 against every other model

Every head-to-head featuring Kimi K3. Verdicts shown for scored pairs.

Kimi K3 vs Fusion
Fusion leads 31–15
Kimi K3 vs Claude Opus 5
Claude Opus 5 leads 22–21
Kimi K3 vs Hermes MoA
Kimi K3 leads 23–21
Kimi K3 vs GPT-5.6 Sol
GPT-5.6 Sol leads 21–20
Kimi K3 vs Claude Fable 5
Kimi K3 leads 29–17
Kimi K3 vs Qwen 3.8
Qwen 3.8 leads 22–17
Kimi K3 vs Grok
Kimi K3 leads 25–18
Kimi K3 vs MiniMax M3
Kimi K3 leads 30–16
Kimi K3 vs Fugu Ultra
Kimi K3 leads 27–15
Kimi K3 vs GLM-5.2
Kimi K3 leads 30–14
Kimi K3 vs Fugu Mini
Kimi K3 leads 25–10
Kimi K3 vs Muse Spark 1.2
Kimi K3 leads 33–13
Kimi K3 vs Opus 4.8
Kimi K3 leads 33–13
Kimi K3 vs Kimi K2.7
Kimi K3 leads 16–9
Kimi K3 vs Qwable 5 27B Coder
Kimi K3 leads 33–8
Kimi K3 vs Gemini 3.6 Flash
Kimi K3 leads 37–11
Kimi K3 vs Claude Sonnet 5
Kimi K3 leads 35–9
Kimi K3 vs Qwen 3.7
Kimi K3 leads 37–10
Kimi K3 vs Fugu Ultra 1.1
Kimi K3 leads 13–8
Kimi K3 vs Inkling
Kimi K3 leads 45–5
Kimi K3 vs Grok 4.6
Kimi K3 leads 16–3
Kimi K3 vs Agents-A1
Kimi K3 leads 42–3
Kimi K3 vs Gemma 4 12B · MLX
Kimi K3 leads 40–2
Kimi K3 vs Laguna XS 2.1
Kimi K3 leads 41–1
Kimi K3 vs Qwythos 9B
Kimi K3 leads 42–0
Kimi K3 vs LongCat-2.0
Tied 2–2
Kimi K3 vs Hy3
Kimi K3 leads 6–1
Kimi K3 vs Gemma-4 12B Coder
Kimi K3 leads 5–1
Kimi K3 vs DeepSeek V4 Flash
50 shared tasks · unscored
Kimi K3 vs DeepSeek V4 Pro
50 shared tasks · unscored
Kimi K3 vs Kimi K2.7 · Fast
47 shared tasks · unscored
Kimi K3 vs Kimi K2.7 · No-Think
47 shared tasks · unscored
Kimi K3 vs Kimi K2.7 · Quality
47 shared tasks · unscored
Kimi K3 vs Ornith 1.0
42 shared tasks · unscored
Kimi K3 vs Claude Mythos 5
Reference-only
Kimi K3 vs Kilo Code
Reference-only

See all 66 comparisons across every model →

Quick pill index

Direct comparisons against every other scored model on the bench:

Kimi K3 vs Fusion Kimi K3 vs Claude Opus 5 Kimi K3 vs Hermes MoA Kimi K3 vs GPT-5.6 Sol Kimi K3 vs Claude Fable 5 Kimi K3 vs Qwen 3.8 Kimi K3 vs Grok Kimi K3 vs MiniMax M3 Kimi K3 vs Fugu Ultra Kimi K3 vs GLM-5.2 Kimi K3 vs Fugu Mini Kimi K3 vs Muse Spark 1.2 Kimi K3 vs Opus 4.8 Kimi K3 vs Kimi K2.7 Kimi K3 vs Qwable 5 27B Coder Kimi K3 vs Gemini 3.6 Flash Kimi K3 vs Claude Sonnet 5 Kimi K3 vs Qwen 3.7 Kimi K3 vs Fugu Ultra 1.1 Kimi K3 vs Inkling Kimi K3 vs Grok 4.6 Kimi K3 vs Agents-A1 Kimi K3 vs Gemma 4 12B · MLX Kimi K3 vs Laguna XS 2.1 Kimi K3 vs Qwythos 9B Kimi K3 vs LongCat-2.0 Kimi K3 vs Hy3 Kimi K3 vs Gemma-4 12B Coder

Read more on agentos.guide: /kimi-k3-machine

Kimi K3 — frequently asked

What is Kimi K3?

Kimi K3 is Moonshot AI's AI model — Moonshot's 2.8T flagship — 1M context, tuned for long-horizon agent work. It has a 1M tokens context window and was released 2026-07-16.

How good is Kimi K3 at coding and one-shot builds?

On the GoldieBench one-shot build benchmark it averages 7.89/10 across 50 scored tasks, with 8 gold, 4 silver and 7 bronze medals.

How much does Kimi K3 cost?

$3 / M in. Launched July 16, 2026. 2.8T-param MoE (Moonshot's quickstart corrected the circulating 2.5T estimate). $3/M input on OpenRouter at launch; included at no extra cost in the Kimi coding plan (`k3` on the coding endpoint).

Where can I see Kimi K3 demos?

Every one-shot build is live and playable on this page and on the GoldieBench compare matrix — same prompt as every other model, no retries.

The same stack Julian uses

Run this stack yourself.

Every demo on this bench was built inside the Agent Operating System — one prompt, one shot, single HTML file out. The Agent OS, the prompts, the templates, the weekly walkthroughs and 4,000+ founders shipping with it every day all live inside the AI Profit Boardroom.

4,000+founders
258documented wins
38countries
$59/momonthly