Microsoft MAI-Image-2.6
B Tier · 7.4/10
Microsoft's in-house image model -- **MAI-Image-2.6 launched 2026-08-10: No. 2 on Arena text-to-image (+79 Elo over 2.5, text rendering +91), and as of 2026-09-04 back to **No. 2 on Arena image editing** as well, plus No. 2 text-to-image and No. 1 image editing on Artificial Analysis -- ahead of Google's Nano Banana family and Meta's Muse Image. **On 2026-09-04 it reached Microsoft Foundry for developers alongside a new MAI-Image-2.6-Flash** (2.8x faster than GPT-Image-2-Medium at 72% greater efficiency), both adding multi-image reference editing, web grounding and dynamic aspect ratios.** The prior MAI-Image-2.5 launched 2026-06-02 at Build (#2 image editing, #3 text-to-image on Arena, surpassing Nano Banana Pro's Arena score; +75 overall vs MAI-Image-2). Live in PowerPoint + OneDrive, on Azure Foundry, MAI Playground, and OpenRouter. Ultra-efficient MAI-Image-2.5-Flash now LIVE in Foundry ($1.75/$1.75/$19.50 per 1M). Original MAI-Image-2 shipped 2026-04-02
Score Breakdown
The Good and the Bad
What we like
- +Debuted #3 on the Arena.ai image model families leaderboard at launch -- a genuinely competitive result against Nano Banana 2, Midjourney, and Flux without Microsoft having shipped an image model before 2026
- +32K-token text input means richer prompts than Nano Banana 2's standard input window -- good for detailed commercial design briefs and multi-element compositions
- +Azure Foundry native -- Microsoft enterprise customers get a first-party image option without an OpenAI dependency, same pattern as MAI-Voice-1 and MAI-Transcribe-1
- +MAI-Image-2-Efficient (2026-04-14 variant) is 22% faster and 4x more efficient -- makes high-volume use cases (batch ad creative, programmatic imagery) materially cheaper without changing the architecture
What could be better
- −Photorealism-first diffusion approach. **The 'Nano Banana wins on text rendering' line no longer holds as of MAI-Image-2.6 (2026-08-10): text rendering gained +91 Elo on text-to-image and +43 on editing, and Microsoft places 2.6 ahead of the Nano Banana family on the Arena editing board.** Midjourney still wins on stylized artistic output. Flux still wins on fine-grained open-source control
- −Not yet available as a consumer web tool -- Bing Image Creator is the closest consumer surface but it has its own UX constraints and limits
- −Azure Foundry token-based pricing ($33/M image output tokens) requires computing effective per-image cost at your resolution. Comparing directly to Nano Banana 2's $0.067/image at 1K is not one-to-one
- −Microsoft has not yet shipped an equivalent of Nano Banana 2's multi-image reference mode, which is the most-requested feature for brand-consistent commercial work
Pricing
MAI-Image-2.6 (2026-08-10) -- MAI Playground + PRIVATE PREVIEW on Foundry
- ✓**Microsoft has published NO rate card for 2.6** -- the launch post states only 'available in MAI Playground and in Private Preview on Microsoft Foundry'. Do not assume it inherits 2.5's $5/$8/$47
- ✓No. 2 on the Arena text-to-image leaderboard at launch; **+79 Elo over MAI-Image-2.5** overall, with gains in every text-to-image category
- ✓**Text rendering +91 Elo** vs 2.5 -- the single largest category gain
- ✓Update 2026-08-18: **No. 3 on Arena image editing**, up from No. 5, **+19 points** vs 2.5 -- biggest gains Text Rendering (+43) and Product, Branding & Commercial Design (+38)
- ✓Access is gated -- Private Preview means request-based, not self-serve
MAI-Image-2.5 (Azure Foundry / OpenRouter, launched 2026-06-02)
- ✓Text input: $5/1M tokens
- ✓Image input: $8/1M tokens
- ✓Image output: $47/1M tokens
- ✓#2 image editing + #3 text-to-image on Arena; surpasses Nano Banana Pro's Arena score
MAI-Image-2.5-Flash (LIVE)
- ✓Ultra-efficient variant of MAI-Image-2.5
- ✓~2.4x cheaper image output than full 2.5
- ✓For high-volume / latency-sensitive workloads
- ✓NOW AVAILABLE in Foundry (vendor post: 'now available to developers in Foundry today', verified 2026-06-09) + OpenRouter
MAI-Image-2 (original, 2026-04-02) + Efficient variant
- ✓Debuted #3 on Arena.ai image families at launch
- ✓MAI-Image-2-Efficient (2026-04-14): 22% faster, 4x more compute-efficient
- ✓Global standard deployment in US + West Europe + Sweden Central + South India
Bundled (Copilot / Bing Image Creator / PowerPoint)
- ✓Existing Microsoft 365 Copilot subscriptions use MAI-Image-2 under the hood
- ✓Bing Image Creator is the consumer-facing surface
- ✓No separate pricing or config required for existing Microsoft customers
Known Issues
- MAI-IMAGE-2.6 REACHES MICROSOFT FOUNDRY, AND A NEW MAI-IMAGE-2.6-FLASH JOINS IT (2026-09-04, vendor-primary): Microsoft published a **new, separately dated post** moving MAI-Image-2.6 out of the private preview this page recorded on 2026-08-10: '**Today, we're bringing it to developers in Microsoft Foundry and expanding the family with MAI-Image-2.6-Flash -- built to deliver that same level of quality for latency-sensitive, high-throughput production workloads.**' **This closes the 'no public rate card, private preview only' caveat that has sat on this page since August.** **NEW CAPABILITIES ON BOTH MODELS:** multi-image reference editing, web grounding, and dynamic aspect ratios. Microsoft frames the pair as a precision-versus-throughput choice: 'maximum precision with MAI-Image-2.6 and production speed with MAI-Image-2.6-Flash.' **MAI-IMAGE-2.6-FLASH PERFORMANCE, VENDOR-PUBLISHED:** '**2.8x faster than GPT-Image-2-Medium while delivering 72% greater efficiency**', at comparable quality. **THE RANK MOVED AGAIN, AND IT MOVED UP -- WHICH IS WHY THIS PAGE DATES ITS RANKS.** As of this post, MAI-Image-2.6 ranks **No. 2 on Arena for BOTH text-to-image and image editing** (as of 2026-09-04), and on **Artificial Analysis** it is **No. 2 for text-to-image and No. 1 for image editing**. **Track the editing rank across this page's own history, because it is a lesson in how fast these decay: 2.5 was No. 2 in June, had silently slipped to No. 5 by mid-August, 2.6 recovered it to No. 3 on 2026-08-18, and it is No. 2 as of 2026-09-04.** Treat every one of those as a dated snapshot, not a standing fact. Microsoft claims 'the best price-per-Elo performance in the world' -- a framing that is quality-per-dollar, not absolute quality, and should not be read as a claim to the top of either leaderboard. **A SOURCING NOTE FOR ANYONE RE-CHECKING THIS:** the URL that previously served the 2026-08-10 launch post now serves this 2026-09-04 post, with `datePublished` and `dateModified` both set to 2026-09-04 and no 'Updated as of' block. The August 18 editing-rank update recorded on this page is **no longer visible on the live page** -- it is preserved here rather than in Microsoft's copy.Source: Microsoft AI (microsoft.ai/news/pushing-the-quality-cost-frontier-with-mai-image-2-6/, JSON-LD datePublished 2026-09-04) -- fetched 2026-09-05 via curl with browser UA; blog index enumerated via Firecrawl · 2026-09-04
- STALENESS CATCH -- MAI-IMAGE-2.6 SHIPPED 2026-08-10 AND THIS PAGE HAD NEVER MENTIONED IT (recorded 2026-08-20, vendor-primary): a ten-day gap that three consecutive sweeps missed, because the vendor is Microsoft AI (microsoft.ai/news) rather than one of the newsrooms enumerated by default. **WHAT LAUNCHED:** MAI-Image-2.6, announced **2026-08-10**, '**ranked second on the Arena text-to-image leaderboard**', improving '**+79 Elo over MAI-Image-2.5 overall, with gains across every Arena text-to-image category**'. **Text rendering alone improves by +91 Elo.** Microsoft's own framing: the result 'firmly establishes MAI-Image ahead of leading models from Meta, Google and xAI'. **THE 2026-08-18 UPDATE TO THE SAME POST:** 2.6 also reached '**No. 3 on Arena's image editing leaderboard, ahead of Google's Nano Banana family and Meta's Muse Image**', a **+19 point gain over 2.5**, climbing **from No. 5 to No. 3**, with its biggest improvements in **Text Rendering (+43)** and **Product, Branding & Commercial Design (+38)**. **A DETAIL WORTH READING CAREFULLY RATHER THAN CELEBRATING:** this page previously described MAI-Image-2.5 as '**#2 image editing**' -- a June snapshot. The 8/18 update reveals 2.5 had **slipped to No. 5 on editing** before 2.6 pulled the family back to No. 3. So the leaderboard position in our old copy had silently decayed, and **2.6 on editing (No. 3) is a recovery to below where 2.5 originally sat, not a new high**. Text-to-image is the genuine improvement: No. 3 to No. 2. **AVAILABILITY IS NARROWER THAN 2.5's:** '**MAI-Image-2.6 is now available in MAI Playground and in Private Preview on Microsoft Foundry**' -- so unlike 2.5 it is **not** generally available on Foundry, and there is no mention of OpenRouter, PowerPoint, OneDrive or Bing Image Creator carrying it. **NO PRICING IS PUBLISHED ANYWHERE IN THE POST** -- treat 2.5's $5/$8/$47 as applying to 2.5 only. **BENCHMARK CAVEAT:** every number here is Arena Elo as reported by Microsoft in its own launch post; Arena is a third-party leaderboard, but the deltas and category breakdowns are Microsoft's presentation of it and no independent write-up is citedSource: Microsoft AI (microsoft.ai/news/mai-image-2-6-launches-at-no-2-on-arena-ahead-of-google-meta-and-xai/, dated August 10, 2026, carrying an 'Updated as of August 18th, 2026' block, fetched 2026-08-20 via curl) · 2026-08-10
- VERSION BUMP (2026-06-02, Microsoft Build): MAI-Image-2.5 launched as part of the 'seven new MAI models' wave. Vendor-published Arena results: #2 for image editing and #3 for text-to-image, with an overall Arena score 'surpassing Nano Banana Pro' and +75 points over MAI-Image-2 (text rendering +107, cartoon/anime/fantasy +90). New capabilities: advanced visual reasoning for scene understanding, fine-grained localized edit control, and face/identity consistency preservation across edits. Now live in PowerPoint and OneDrive (presentation visuals + photo editing), on Azure Foundry, MAI Playground, and via OpenRouter. Pricing: $5/1M text, $8/1M image-in, $47/1M image-out. The ultra-efficient MAI-Image-2.5-Flash ($1.75/$1.75/$19.50 per 1M) is NOW LIVE in Foundry alongside it (verified vendor-primary 2026-06-09).Source: Microsoft AI (microsoft.ai/news/introducing-mai-image-2-5/, microsoft.ai/news/building-a-hillclimbing-machine-launching-seven-new-mai-models/) · 2026-06-02
- Public preview on Azure Foundry -- availability is region-dependent. Global Standard deployment covers US + West Europe + Sweden Central + South India at launch. Other regions need to fall back to nearest availableSource: Microsoft Foundry catalog, Microsoft AI blog · 2026-04
- Model card dated 2026-03-18 internally, publicly announced 2026-04-02 -- Microsoft has been running the model internally for several weeks before opening public preview, which explains the scale of Copilot/Bing integration at launchSource: Microsoft model card PDF · 2026-04
Best for
Microsoft shops already on Azure or M365 Copilot who need a first-party image model without an OpenAI dependency. Also good for any high-volume programmatic image workflow (ad creative, product photography variations) where MAI-Image-2-Efficient's 4x cost efficiency materially changes the economics.
Not for
Text-heavy commercial design (use Nano Banana 2). Stylized artistic work (use Midjourney). Open-weight self-hosting requirements (use FLUX.2 [klein]). Consumer creators who want a simple web UI -- the Foundry workflow is developer-facing.
Our Verdict
MAI-Image-2.5 (2026-06-02) is the version bump that closes the gap. Where MAI-Image-2 debuted #3 on Arena as a surprise, 2.5 now ranks #2 for image editing and #3 for text-to-image, with an overall Arena score Microsoft says surpasses Nano Banana Pro -- and the +107-point text-rendering jump directly targets the one area where Nano Banana used to win. The new fine-grained localized editing and face/identity consistency make it genuinely competitive for commercial work, and being live in PowerPoint + OneDrive means most Microsoft 365 users already have it. For Azure customers this is now a first-tier image option, not just a no-OpenAI-dependency fallback. Midjourney still wins on stylized art and Flux on open-weight control, but the gap is much smaller than it was in April. The cheaper MAI-Image-2.5-Flash is now live in Foundry for high-volume use.
Sources
- Microsoft AI: Pushing the quality-cost frontier with MAI-Image-2.6 (2026-09-04) -- Foundry GA + MAI-Image-2.6-Flash, Arena No. 2 text-to-image and editing (accessed 2026-09-05)
- Microsoft AI: MAI-Image-2.6 launches at No. 2 on Arena, ahead of Google, Meta and xAI (2026-08-10, updated 2026-08-18) (accessed 2026-08-20)
- Microsoft AI: Introducing MAI-Image-2.5 (2026-06-02) (accessed 2026-06-02)
- Microsoft AI: Launching seven new MAI models (2026-06-02) (accessed 2026-06-02)
- Microsoft AI: 3 new MAI models in Foundry (accessed 2026-04-17)
- Microsoft Foundry model catalog: MAI-Image-2 (accessed 2026-04-17)
- Microsoft Community Hub: MAI-Image-2-Efficient (accessed 2026-04-17)
- Microsoft Learn: Foundry Models docs (accessed 2026-04-17)
Explore more Microsoft MAI-Image-2.6 rankings
Deeper leaderboards, benchmarks, task-specific tier lists, and status/pricing pages for Microsoft MAI-Image-2.6.
The Tier List Tuesday
Weekly newsletter: tier movers, new entrants, and the VS of the week. Built from our daily AI-tool sweeps. No spam, unsubscribe anytime.
Alternatives to Microsoft MAI-Image-2.6
Midjourney
Industry-leading AI image generation with stunning artistic quality. V8.2 launched 2026-07-24 as the new default model -- an aesthetics and quality pass that Midjourney says makes output 'more creative, bold, sophisticated, edgy and fresh' while dramatically cutting low-quality results, plus better Personalization for accounts with long rating histories. No price change
DALL-E (Shut Down)
OpenAI's DALL-E 2 and DALL-E 3 -- SHUT DOWN. Both APIs were retired on 2026-05-12. DALL-E 3 was already removed from ChatGPT in December 2025. Existing integrations now fail; migrate to gpt-image-1.5 / gpt-image-1-mini (request shape differs -- not a drop-in swap). Tier-list alternatives: Nano Banana 2, Midjourney, FLUX.2 [klein], Ideogram
Stable Diffusion
Open-source AI image generation with unlimited free local use and full customization
Leonardo AI
Versatile AI image generator with fine-tuned models and a generous free tier
Adobe Firefly
Adobe's AI image generator -- commercially safe and baked into Creative Cloud. Firefly AI Assistant public beta is now LIVE globally (2026-04-27) for Creative Cloud Pro and paid Firefly plans, orchestrating multi-step workflows across Photoshop, Lightroom, Premiere, and Illustrator from a single chat
Ideogram
AI image generator that nails text rendering -- Ideogram 4.0 (June 3 2026) goes OPEN-WEIGHT with commercial license: 9.3B DiT, 0.97 X-Omni OCR, layout control, 2K photoreal. Plus Custom Models for brand-trained generation on Pro+ plans
![Flux (FLUX.2 [klein]) logo](/logos/flux.png){kind=small}
Flux (FLUX.2 [klein])
Black Forest Labs open-source image model -- FLUX.2 [klein] (Jan 15 2026) is the fastest image model to date at sub-0.5s generation, 4MP coherence, multi-reference, and native editing. 4B + 9B open-core variants
Krea AI
Real-time AI image generation and enhancement with a visual, interactive canvas
NightCafe
Community-driven AI art generator with multiple models, daily free credits, and a social gallery
Nano Banana 2 (Gemini 3.1 Flash Image)
Google's Gemini 3.1 Flash Image model -- the best-in-class text-in-image renderer, now the default across the Gemini app
Meta Muse (Image + Video)
Meta Superintelligence Labs' first media-generation models (launched 2026-07-07) -- Muse Image is GA across the Meta AI app, Instagram, and WhatsApp (Arena #2 for text-to-image); Muse Video is in preview (Arena #3 for text-to-video). Instruction-faithful editing, multi-reference composition, tool use, and Content Seal provenance watermarking