MuscleWiki API vs ExerciseDB (AscendAPI)

ExerciseDB now trades as AscendAPI. It advertises the largest exercise catalogue of the vendors compared here, and it sells its paid tiers through the RapidAPI marketplace.

Sources: AscendAPI docs: ExerciseDB V2 overview, AscendAPI docs: authentication

The MuscleWiki API is built around the demonstration: 7,766 MP4 clips across 1,943 exercises, male and female, front and side, each with a matching thumbnail. ExerciseDB's own documentation says each of their exercises includes one MP4 demonstration video. We also ship 14 languages at parity against the single language their translation table marks Available, and media URLs that never rotate. ExerciseDB advertises the bigger catalogue, 11,000+ exercises against our 1,943, so a product that needs the long tail of machine variants will find more of it there. If what your users watch is the demonstration, ours has more of it in every record.

Sources: AscendAPI docs: ExerciseDB V2 overview, AscendAPI docs: ExerciseDB V1 overview, AscendAPI docs: translation status table, AscendAPI docs: caching policy

Every claim about another vendor on this page was fetched from that vendor’s own public documentation. This one was verified 14 September 2026. Figures change; if you are reading this long after that date, check the linked sources before you rely on it.

Side by side

Every figure in the ExerciseDB column links to the page we read it on. Where we could not fetch a fact from a public source, there is no row for it.

The MuscleWiki API and ExerciseDB (AscendAPI) compared across 15 axes, with a source link for every claim about ExerciseDB.
AxisMuscleWiki APIExerciseDB
Exercises in the catalogue1,943 (counted from the data we serve, 10 September 2026)V2: "11,000+ structured exercises". V1: "over 2,000 structured exercises", of which 1,500 are in the free tierSources: AscendAPI docs: ExerciseDB V2 overview, AscendAPI docs: ExerciseDB V1 overview, ExerciseDB API repository: README and licence
Video per exercise7,766 MP4 clips: 4 for every exercise, male and female, front and side. Every clip carries a matching thumbnailV2: "Each exercise includes an MP4 demonstration video". One clip, one angle. V1 ships GIF animations rather than video (180p on the free tier)Sources: AscendAPI docs: ExerciseDB V2 overview, AscendAPI docs: ExerciseDB V1 overview
Branding on the videoMuscleWiki branding is embedded in every clip on every plan, and our Terms require that it stays visible. Upgrading does not remove it; the assets are identical across plansV2: "Free tier media is watermarked. Paid tier serves clean, watermark-free assets across all resolutions"Source: AscendAPI docs: ExerciseDB V2 overview
Still imagesOne thumbnail per video (7,766), plus 15,144 bodymap images highlighting the muscles a movement works (Professional plan or higher)V2: "20,000+ exercise images" at 360p/480p/720p/1080p. Their README also advertises "5,000+ GIF animations" across the productsSources: AscendAPI docs: ExerciseDB V2 overview, ExerciseDB API repository: README and licence
Languages14 languages at parity: every exercise carries its name in all 14 and returns the same instruction steps in all 14, so no language returns a thinner record than English (non-English needs the Growth plan or higher)Their published translation status table lists one language as Available (en-GB), German and Spanish as In Progress, and 38 more as PlannedSource: AscendAPI docs: translation status table
Routines and workouts32 curated routines and 92 workouts as read-only API resources, with sets, reps and exercise references (Professional plan or higher)Their documented endpoints cover exercises, body parts, muscles, equipment, exercise types and muscle visualisation, and the published page index lists no routine or workout resource today. A separate WorkoutDB product is carded "Coming Soon", with "full training plans, sets, reps, rest periods, and progressions"Sources: AscendAPI docs: index of every published page, AscendAPI docs: product introduction and roadmap cards
Difficulty metadataPresent on every exercise (Novice, Beginner, Intermediate, Advanced)V2 lists difficultyLevel in its field table with the status "Coming Soon". V1's paid tier advertises difficulty levels todaySources: AscendAPI docs: ExerciseDB V2 overview, AscendAPI docs: ExerciseDB V1 overview
How you buy itDirect from us: create an account and generate a key in your dashboardThrough the RapidAPI marketplace for both the free and paid V2 tiers. The V1 free endpoint is hosted directly and needs no accountSources: AscendAPI docs: ExerciseDB V2 overview, AscendAPI docs: ExerciseDB V1 overview
AuthenticationA single X-API-Key header on every request. Your server can exchange that key for a short-lived media token, so a native player fetches /stream/ URLs with no header and the key never ships inside an app binary"All AscendAPI endpoints are authenticated via your RapidAPI key": X-RapidAPI-Key plus an X-RapidAPI-Host headerSources: AscendAPI docs: authentication, AscendAPI docs: ExerciseDB V2 overview
Media URL stability and cachingStable, deterministic media URLs. Caching allowance: text metadata up to 30 days, thumbnails 24 hours on-device only, video and bodymaps served no-storeMedia asset URLs "rotate every week on Monday at 00:00 UTC", and caching "is only permitted if your current plan explicitly allows it"Source: AscendAPI docs: caching policy
Rate limitingA monthly call quota per plan rather than a per-second limit; over-quota requests return 429 with Retry-AfterFree plans: "1,000 requests per hour" per API key. Paid plans: "no strict rate limit"Source: AscendAPI docs: rate limiting
Free tier500 calls a month in the Playground on this site. A Basic account does get a key, but it works only through the Playground: the API refuses it without a valid Playground signature. Direct access starts at $10 a monthV1: 1,500 exercises with 180p GIFs, no sign-up and no key. V2: a free RapidAPI tier whose media is watermarkedSources: AscendAPI docs: ExerciseDB V1 overview, AscendAPI docs: ExerciseDB V2 overview
Guidance on their exploration endpointsOne API, documented for production useTheir README says the v2 playground endpoints are "for exploration only and not recommended for production integration", citing "strict rate limits and potential instability"Source: ExerciseDB API repository: README and licence
Code licenceNot open source; a hosted service with commercial termsThe GitHub repository their documentation links is marked AGPL-3.0, but publishes only a LICENSE and a README, with no source codeSource: ExerciseDB API repository: README and licence
MCP server for AI agentsFirst-party MCP server, npx @musclewiki/mcp, which forwards every call to the API using your own keyNone in their published documentation index today; their introduction announces an "upcoming MCP Server integration"Sources: AscendAPI docs: index of every published page, AscendAPI docs: product introduction and roadmap cards

Sources for our column

Exercise, video, language and bodymap counts all come from the dataset the API serves. Every figure in the ExerciseDB column links to the page it came from, in the cell that makes the claim.

Where the MuscleWiki API and ExerciseDB differ

The handful of differences that actually decide it, and what each one costs whoever picks the other side.

Count the rows, or look inside one
ExerciseDB V2 advertises "11,000+ structured exercises", roughly 5.7 times our 1,943. Their documentation is just as clear about what sits inside one: "Each exercise includes an MP4 demonstration video". One clip, one angle. Ours carry male and female footage, front and side, 7,766 clips in total. Plenty of products need the breadth more than the depth. A headline count will not tell you which yours is.Sources: AscendAPI docs: ExerciseDB V2 overview, ExerciseDB API repository: README and licence
Their media URLs rotate every week
AscendAPI's caching guide says media asset URLs "rotate every week on Monday at 00:00 UTC", and that caching "is only permitted if your current plan explicitly allows it". Ours have no signing, no expiry and no rotation schedule, so there is no weekly re-fetch job to write. A stored URL changes only when that exercise's media does, which is a catalogue update rather than a clock.Source: AscendAPI docs: caching policy
What their watermark is for
Their V2 overview says "Free tier media is watermarked. Paid tier serves clean, watermark-free assets across all resolutions". The mark is a paywall device, so what you evaluate free is not what you would ship. MuscleWiki branding is embedded in every clip on every plan and our Terms require it to stay visible, which cuts both ways: you cannot buy it off, and the clips in our Playground are the files a paid key returns.Sources: AscendAPI docs: ExerciseDB V2 overview, MuscleWiki API Terms: §8 Attribution
Buying through RapidAPI
Their documentation says "All AscendAPI endpoints are authenticated via your RapidAPI key", sent alongside an X-RapidAPI-Host header, and their paid V2 tiers are sold through that marketplace. With us you create an account and send a single X-API-Key. No proxy in the path, no marketplace terms on top of ours.Sources: AscendAPI docs: authentication, AscendAPI docs: ExerciseDB V2 overview
Four resolutions of still images
ExerciseDB V2 serves static images at 360p, 480p, 720p and 1080p, and their README advertises GIF animations too. We serve one thumbnail per clip, plus 15,144 bodymap images showing the muscles a movement works (Professional plan or higher). Their resolution ladder gives a dense mobile grid more to work with. On a detail screen, the video is what the user opened the app for.Sources: AscendAPI docs: ExerciseDB V2 overview, ExerciseDB API repository: README and licence
What their AGPL-3.0 tag actually covers
The GitHub repository their documentation links is marked AGPL-3.0, and it holds two files: a LICENSE and a README. No server code is published there, and the licence notice it carries is applied to the program rather than to the data. Ours is a closed hosted service and says so. On the question a licence tag is usually read to answer, whether you can run the thing yourself and keep the data, neither of us hands that over.Source: ExerciseDB API repository: README and licence

Moving an existing ExerciseDB integration to us

What actually changes in your code.

Swap the auth headers
Replace X-RapidAPI-Key and X-RapidAPI-Host with a single X-API-Key header carrying a key you generate in your dashboard. There is no marketplace in the path, so there is no proxy secret to configure.Sources: AscendAPI docs: authentication, AscendAPI docs: ExerciseDB V2 overview
Map the fields
Their bodyParts and targetMuscles collapse into our primary_muscles. equipments becomes category, instructions becomes steps, and videoUrl becomes a videos array whose entries carry url, angle, gender and og_image. Their exerciseId strings (exr_…) have no equivalent; our IDs are integers.Source: AscendAPI docs: ExerciseDB V2 overview
Delete the weekly media-URL refresh
Their media URLs rotate weekly, so most integrations re-fetch to avoid 404s. Ours have no signing, no expiry and no rotation schedule, so there is no weekly re-fetch job to write. A stored URL changes only when that exercise's media does, which is a catalogue update rather than a clock. Read our caching rules first, though: text metadata up to 30 days, thumbnails 24 hours on-device only, video and bodymaps served no-store.Source: AscendAPI docs: caching policy
Add the lang parameter where you had English-only text
Every data endpoint takes lang, and all 14 languages return the same fields, so a translated response is never a thinner one. If you were running machine translation over an English-only response, that layer comes out. Non-English needs the Growth plan or higher.Source: AscendAPI docs: translation status table

You can try every endpoint in the Playground before you commit, read the documentation, or check what each plan reaches on pricing.

Other comparisons