Best AI music generators for video creators (2026): What ChatGPT, Claude & Gemini Recommend

Suno is the AI music generator the models name most often for video creators. It appears in 45 of 50 recorded answers and comes first in 32. Udio is the next most-named product, and Soundraw is the next most common first pick. The list below ranks 15 products by how often ten AI models named them across five fixed buyer questions in edition 2026-09.

TL;DR

• Suno is the default name. Every model includes it in at least four of its five answers, and it opens 32 of 50.
• Udio is the standing second choice, with 33 mentions and no first positions.
• Background-music tools split the panel. Soundraw leads the tally for Claude Fable 5, Perplexity and Sonar Reasoning Pro.
• The OpenAI models show zero Soundraw mentions only because they write the name as SOUNDRAW. Its row is a floor.
• Freebeat and Neural Frames are music-video tools. They sit low in the count because they solve a different job.
• Choose by job first. Then read the commercial terms of the paid plan before any client work.

1. Suno

Pick Suno if the video needs a finished song with vocals and you want the name every model on the panel reaches for.

Measured: named 45 of 50 (Suno 90%), named first 32 of 50 (Suno 64%), average position 1.64.

Suno is a web platform for making, editing, remixing and sharing songs. Users start from a prompt or from its editing tools. They can upload or record audio, rewrite lyrics and reorder sections. The split shows how settled the models are. All ten name Suno in four or five of their five answers, so no provider leaves it out. Its soft spot is background music. Claude Fable 5, Perplexity and Sonar Reasoning Pro each name Suno four times and Soundraw five times. Their top name is a background-music tool, and Suno is built around songs.

Pros

• Named by all ten models, at four or five answers each
• Opens 32 of the 50 answers, the highest first count in the edition
• Paid subscribers keep the songs they create and can use them as they wish
• Studio editing brings DAW-style tools into the same web app as generation

Cons

• Free-plan tracks cannot be used commercially, per Somio’s guide
• Leader for seven of the ten models, with Soundraw ahead for the other three

Pricing: a free plan, plus Pro and Premier paid plans. Best for: anyone who wants to make music, from first-time creators to experienced musicians.

2. Udio

Pick Udio as the second name on a shortlist if you want to rework a song section by section after the first generation.

Measured: named 33 of 50 (Udio 66%), named first 0 of 50 (Udio 0%), average position 3.21.

Udio turns text descriptions into songs. Its editing tools extend, inpaint and remix tracks, including audio you upload. Finished songs can be published on Udio, where users follow other artists. The counts give it a fixed role. Udio is the second most-named product in the edition and the first name in none of its 33 answers. Its named-versus-first gap of 66 points is the widest in the table. The models list it as the alternative to Suno and never as the lead. Coverage is broad and even. Every model names it, and none names it more than four times in five. Somio’s guide separates the two by job and calls Udio best for precise song refinement.

Pros

• Gemini and Gemini 3.5 Flash name it in four of five answers each
• Inpainting replaces one section of a song and leaves the rest untouched
• Every model names it at least twice

Cons

• Zero first positions across 50 answers
• Export is not always straightforward on some plans, per Somio’s guide
• Only two of five GPT-5.6 Sol answers mention it

Pricing: a free plan, Standard and Pro plans billed monthly or annually, and purchasable credits. Best for: professional musicians, songwriters and amateurs who want to edit what the model produces.

3. Soundraw

Pick Soundraw if the music sits under narration and has to fit the length and energy of your edit.

Measured: named 28 of 50 (Soundraw 56%), named first 11 of 50 (Soundraw 22%), average position 2.29.

Soundraw combines chosen genres into royalty-free tracks. Its Mixer changes instruments, intensity and track length after generation. Depending on the plan, tracks download as MP3, WAV or separate stems. Its model is trained on music written by Soundraw’s own producers. When a model names Soundraw, it tends to name it early. Five models name it in every answer, and three of them name it more often than Suno. The three OpenAI models show zero. Their answers name it repeatedly, in capitals the counter did not match, so read 28 of 50 as a floor.

Pros

• Five models name it in every answer they gave
• Opens 11 of 50 answers, the second-highest first count
• Average position of 2.29, behind only Suno

Cons

• No vocal or lyric generation, per Somio’s guide
• Free users can preview tracks but need a paid plan to download them
• Claude Opus 5 names it in one answer of five

Pricing: monthly and annual paid plans for individuals, with enterprise pricing on request. Best for: content creators, music artists and producers, and companies.

4. Beatoven

Pick Beatoven if the mood has to shift inside one video and you would sooner describe the scene than choose genres.

Measured: named 25 of 50 (Beatoven 50%), named first 3 of 50 (Beatoven 6%), average position 3.24.

Beatoven builds background music from a written description with its maestro model. Its stated uses run from YouTube video to podcasts, games, trailers, ads and livestreams. The total hides a split by provider. Claude Fable 5 and Sonar Reasoning Pro name it in every answer, and the Perplexity family names it at Beatoven 90%. The Anthropic family sits at Beatoven 66.7%. The OpenAI family drops to Beatoven 26.7%, with no mention at all from ChatGPT. A buyer who asks a Perplexity model will hear about Beatoven. A buyer who asks ChatGPT may never see the name.

Pros

• Third-highest first count in the edition, at 3 of 50
• Different moods can be set for different sections of one track, per Somio’s guide
• Generates sound effects from a prompt as well as music

Cons

• Each Google model names it only once
• Downloads are billed by duration, which adds up on long-form content

Pricing: pay per track, or subscribe for additional download minutes. Best for: creators making video, podcasts, games, short films, ads and livestreams.

5. AIVA

Pick AIVA for orchestral and cinematic scoring, especially if you want to own the copyright outright.

Measured: named 25 of 50 (AIVA 50%), named first 0 of 50 (AIVA 0%), average position 4.2.

AIVA calls itself a music generation assistant. Users can build their own style models and upload audio or MIDI as an influence. They can then edit generated tracks before download. The models treat AIVA as the specialist they add near the end of a list. It ties Beatoven on mentions, yet its average position of 4.2 is later than any product ranked above it. Gemini’s answer to the target question names it the pick for cinematic and orchestral scoring. The split inside one provider is wide. GPT-5.6 Luna names it in four of five answers. GPT-5.6 Sol names it in none.

Pros

• Pro subscribers own the full copyright of their compositions
• Nine of the ten models name it
• Standard plan targets creators monetising on YouTube, Twitch, TikTok and Instagram

Cons

• Never named first in 50 answers
• No AI vocal or lyric generation, per Somio’s guide

Pricing: a free plan, plus Standard and Pro annual plans priced in euros plus VAT. Best for: beginners, music professionals and content creators who want to monetise compositions.

6. Mubert

Pick Mubert if you need a steady stream of background audio, or music generated inside your own app.

Measured: named 24 of 50 (Mubert 48%), named first 1 of 50 (Mubert 2%), average position 4.13.

Mubert generates royalty-free soundtracks from samples supplied by artists. Its product line splits by user. Render serves content creators and an API serves developers and brands. Studio pays artists who contribute samples, and Play streams music to listeners. The streaming and developer side shapes how the models use it. Sonar Reasoning Pro names it in all five answers. Its royalty-free answer puts Soundraw first, “with Mubert as the best alternative if you want adaptive/streaming-style background audio rather than fixed tracks.” Mubert is a Perplexity-family answer first, at Mubert 80% of that family’s answers.

Pros

• An API puts soundtrack generation inside apps and games
• Mubert Render matches a track to a video’s mood and duration

Cons

• Just one of 50 answers opens with it
• Once each from GPT-5.6 Sol, ChatGPT and Gemini, far below its Perplexity rate

Pricing: no public pricing is recorded. Best for: content creators, artists, developers and brands.

7. ElevenLabs

Pick ElevenLabs if music is one output among voiceover, image and video, and you want all of them from one supplier.

Measured: named 14 of 50 (ElevenLabs 28%), named first 0 of 50 (ElevenLabs 0%), average position 4.36.

ElevenLabs runs three platforms, for customer-experience agents, creative media and developers. Music is one product inside the creative suite. The counts match that position. The models name it widely and thinly, and never first. Somio’s guide rates its music tool for realistic vocals and simple text-to-song creation. Part of its thin count is spelling. Claude Fable 5 writes the product as ElevenMusic, and the tracked names for ElevenLabs do not include that form. Its zero from that model understates what that model’s answers say. Treat the row as a floor, as with Soundraw.

Pros

• Music sits beside speech, image and video generation in ElevenCreative
• Eight of the ten models name it

Cons

• Absent from every ChatGPT answer
• MP3 is the main export format, per Somio’s guide
• Never more than three of five answers from any model

Pricing: no public pricing is recorded. Best for: businesses, creators, marketers and developers.

8. Stable Audio

Pick Stable Audio if you produce in a DAW and want generated music and sound effects inside the tools you already use.

Measured: named 8 of 50 (Stable Audio 16%), named first 0 of 50 (Stable Audio 0%), average position 3.88.

Stable Audio creates music and sound effects with conversational controls in a web app. It edits with audio-to-audio and inpainting. A plugin brings it into major DAWs, and an API platform charges per generation. GPT-5.6 Sol’s comparison answer places it for instrumentals and sound design. It appears in few answers, but its average position of 3.88 is earlier than AIVA, Mubert or ElevenLabs manage. GPT-5.6 Sol is the only model to name it twice, and no model opens with it.

Pros

• Plugin runs inside Ableton, Logic, Pro Tools and other major DAWs
• All three OpenAI models name it at least once

Cons

• Stem export is labelled experimental
• Claude, Claude Fable 5 and Sonar Reasoning Pro never mention it

Pricing: API priced per generation, with bulk discounts. Best for: artists and audio professionals.

9. Freebeat

Pick Freebeat if the deliverable is a music video built around a track you already have.

Measured: named 7 of 50 (Freebeat 14%), named first 2 of 50 (Freebeat 4%), average position 4.57.

Freebeat is a music-video generator that turns up in answers about music. Perplexity framed it that way: “I’d recommend Freebeat if your goal is to turn music into a finished music video with minimal manual editing.” A musician’s test on adamharkus.com names it the winner among AI video tools. Renderforest’s comparison rates it best for performance, dance and lip sync. Its reach is narrow. Four models name it, all from Anthropic or Perplexity, and no OpenAI or Google model does. Two of its seven mentions are first positions, more than Udio or AIVA manage across all 50 answers.

Pros

• Analyses BPM, bars and arrangement before it generates scenes
• Claude names it in three of five answers
• Horizontal, vertical and square output are supported

Cons

• Performance shots need careful consistency review, per Renderforest
• Needs a finished track first, from a WAV upload or a Suno link

Pricing: a free entry point and paid credit-based plans, per Renderforest’s comparison. Best for: performance-led videos, AI singers or characters, dance content and lyric videos, per Renderforest.

10. Soundful

Pick Soundful if you want quick instrumental beds built from templates for video and social posts.

Measured: named 7 of 50 (Soundful 14%), named first 1 of 50 (Soundful 2%), average position 3.29.

Soundful has no vendor record in this edition, so its description comes from the recorded answers. Claude describes it as a generator of background music for videos, ads and online content. A user picks a style or mood, generates a track and downloads it as MP3 or WAV, with stems as an option. Its mentions come from one provider. Claude Opus 5 accounts for four, Claude for two and Claude Fable 5 for one. When an Anthropic model reaches for a quick background track, Soundful is one of its answers. A buyer who uses any other assistant is unlikely to hear the name.

Pros

• Templates target YouTube, Instagram, gaming, podcasts and ads, per the recorded answer
• Average position of 3.29, close to Udio’s

Cons

• Silent across every OpenAI, Google and Perplexity answer
• Makes no full songs with vocals or lyrics, per Claude’s answer

Pricing: no public pricing is recorded. Best for: creators who need instrumental background tracks for video, ads and social posts.

11. Loudly

Pick Loudly if you already own a catalogue and want to remix, master and release it, with generation as one tool among several.

Measured: named 5 of 50 (Loudly 10%), named first 0 of 50 (Loudly 0%), average position 4.4.

Loudly generates tracks with its own rights-cleared models and with partner models. It also analyses songs into a fingerprint it calls SongDNA. It remixes songs the user owns, masters uploads and releases music directly to other platforms. Its record addresses musicians, labels, catalogue owners and developers. The counts fit that audience. Five models name it once each, and none names it first. It turns up as a passing mention across providers. The fit is strongest for a buyer who already has music to rework.

Pros

• Stem Splitter separates vocals, drums, bass and more in WAV or MP3
• Mastering exports a polished version guided by platform loudness targets
• Named by models from all four families

Cons

• No model names it more than once
• Opens none of its five answers

Pricing: no public pricing is recorded. Best for: musicians and producers, labels and catalogue owners, and developers.

12. Epidemic Sound

Pick Epidemic Sound if licence certainty matters more to you than generation, because it licenses human-made music and uses AI to fit it to video.

Measured: named 4 of 50 (Epidemic Sound 8%), named first 0 of 50 (Epidemic Sound 0%), average position 4.75.

Epidemic Sound licenses human-made music and sound effects for video, with AI tools for fitting tracks to footage. The answers treat it as a library with AI around it. Gemini’s answer to the target question calls it a hybrid stock library and picks it as the safe choice for YouTube monetisation. Perplexity names it for human-composed music with AI editing and video-friendly licensing. Claude Opus 5 names it twice. Both framings put licensing ahead of generation.

Pros

• Studio syncs human-made music and sound effects to video
• Track suggestions come from frames in your own footage
• Subscriptions cover individuals, professionals, brands and agencies

Cons

• Only three models name it
• Average position of 4.75, later than any product ranked above it

Pricing: Creator, Pro and Business subscriptions billed yearly, plus custom plans. Best for: individual creators, freelancers, brands and agencies making video.

13. Lyria

Pick Lyria only if you are building music generation into your own app, because its record addresses app builders and not video editors.

Measured: named 2 of 50 (Lyria 4%), named first 0 of 50 (Lyria 0%), average position 4.

Lyria is a music-generation model family in Google AI Studio. It turns prompts or visual inspiration into tracks. An API model generates a song from a text prompt. Its record addresses people building custom music apps, and the counts reflect that audience. Neither Google model on the panel names it. Its two mentions come from Claude Opus 5 and Sonar Reasoning Pro. For a video creator, Lyria is relevant only when the music has to be generated inside a product you build.

Pros

• Real-time composition shifts instruments, genres and moods on the fly
• BPM, scale and tone can be set directly for custom music apps

Cons

• Two mentions in 50 answers
• Never the first name in an answer

Pricing: no public pricing is recorded. Best for: people building custom music apps.

14. Neural Frames

Pick Neural Frames if you make electronic or abstract music and want the visuals to react to each part of the track.

Measured: named 2 of 50 (Neural Frames 4%), named first 0 of 50 (Neural Frames 0%), average position 6.5.

Neural Frames is a music-video tool built around audio-reactive visuals. The musician’s test on adamharkus.com says it separates a track into stems and maps visual behaviour to frequency ranges. The same test found no lip-sync. Renderforest’s comparison lists lyric and lip-sync tools in its toolkit. The two guides disagree on a feature that decides performer-led videos, so check the current product first. For abstract electronic work, both guides point to Neural Frames.

Pros

• Renderforest names Neural Frames best for advanced audio-reactive control
• Its own site is among the vendor-owned hosts cited in the answers

Cons

• Arrives late, at an average position of 6.5
• Named once each by only two models, Claude and Perplexity
• No character system for performer-led videos, per the adamharkus.com test

Pricing: credit-based, estimated by song duration and workflow, per Renderforest’s comparison. Best for: electronic, ambient, experimental and instrumental releases, per Renderforest.

15. Boomy

Pick Boomy if you have never made music and want a first original song you can share.

Measured: named 1 of 50 (Boomy 2%), named first 0 of 50 (Boomy 0%), average position 8.

Boomy is built for people who have never made music before. It creates original songs with AI and lets users share them. For a video creator it is a songwriting entry point, and its record says nothing about scoring footage. The count is the thinnest in the edition. Claude names it once, late in the answer, and no other model names it. The mention confirms the models know the product. It says little about fit for video work, so treat Boomy as a starting point for a first song.

Pros

• Full commercial rights come with songs you monetise
• Beginner-first, with no music-making experience assumed

Cons

• One answer out of 50 names it
• Latest average position in the table, at 8

Pricing: no public pricing is recorded. Best for: people making their first song.

How the tools compare

Suno leads on every measure: most mentions, most first positions and the earliest average position.

Below Suno the table falls in two steps. Six products appear in 24 or more answers. ElevenLabs follows at 14, and the last eight sit at 8 or fewer. First positions are more concentrated still. Only Suno and Soundraw open more than three answers, and nine of the fifteen products never open one.

Where the models disagree

The models agree that Suno belongs in the answer and split on the top name. Suno is the most-named product for seven models. Both Google models name Soundraw just as often, in every answer (Suno 100%, Soundraw 100% of Google answers). Claude Fable 5, Perplexity and Sonar Reasoning Pro name Soundraw more often than Suno, each in all five answers.

The families pull apart below the leader. The OpenAI family names Suno in almost every answer (Suno 93.3%), and the Anthropic family is close behind (Suno 86.7%). The Perplexity family names Soundraw in all ten of its answers.

Individual models disagree more sharply. Beatoven appears in every answer from Claude Fable 5 and Sonar Reasoning Pro, and in none from ChatGPT. AIVA runs from four of five with GPT-5.6 Luna to zero with GPT-5.6 Sol, two models from the same provider. Soundful exists only in Anthropic answers. Mubert reaches five of five only with Sonar Reasoning Pro. Freebeat never appears in an OpenAI or Google answer.

Part of the OpenAI split on Soundraw is a counting gap.

Why do the OpenAI models show zero for Soundraw?

The counter missed their spelling. The three OpenAI models are recorded at zero Soundraw mentions, yet their answers name the product again and again, written in capitals as SOUNDRAW.

ChatGPT’s answer to the fourth question opens: “Best default for a video creator: SOUNDRAW.” GPT-5.6 Luna’s answer to the target question reads: “For most video creators, SOUNDRAW is the best choice because it is designed for background music rather than standalone songs.” GPT-5.6 Sol’s answer to the royalty-free question names SOUNDRAW as its first choice.

The panel tracks one name for the product, “Soundraw”. Any tracked name that starts with a capital letter is matched case-sensitively, so ordinary words are not mistaken for vendor names. “SOUNDRAW” does not match “Soundraw”. Each of those answers was counted as silent on Soundraw.

For a buyer, the consequence is specific. Soundraw’s row is a floor. The OpenAI answers that open with it are uncounted first positions, so the true distance between Suno and Soundraw is shorter than the table shows. The ranking stays in measured order, because a recount changes the measurement itself. Read with the raw text, each family has at least one model that treats Soundraw as a lead pick for background music.

How the sample was built

Sample: 10 models x 5 fixed prompts = 50 recorded answers. Each model answered each question once, with web search on. The panel tracked 17 products and named 15. Riffusion and Splash were never named, so they carry no row.

The five recorded questions, verbatim:

1. What is the best AI music generation tool for a video creator? Name specific products.
2. Which AI music generation tool would you recommend to a video creator in 2026?
3. Compare the top AI music generation tool options right now.
4. I’m a video creator and I need an AI music generation tool. What should I use and why?
5. Best AI music generation tool for royalty-free background music?

The ten models come from four families. OpenAI supplies GPT-5.6 Sol, ChatGPT and GPT-5.6 Luna, for 15 answers. Anthropic supplies Claude Opus 5, Claude and Claude Fable 5, for 15. Google supplies Gemini and Gemini 3.5 Flash, for 10. Perplexity supplies Perplexity and Sonar Reasoning Pro, for 10.

Answer share is the share of recorded answers that named a product. Named first is the share of answers where it appeared before any other tracked product. A lower average position means the product arrives earlier in the answers that name it. Every answer and the full per-model split sit in the AI music index. The counting rules are set out in the method.

How this sits against the AI music guides

Most of the readable pages that rank for this question come from vendors. The musician’s test on adamharkus.com covers video tools. None of them counts what AI answers say.

Somio’s guide to AI music generators names Somio, its own product, as the top pick. It reviews Suno, Udio, Mureka, Soundraw, ElevenLabs, Beatoven.ai and AIVA with features, pricing notes, pros and cons. Its stated criteria are music quality, vocal realism, customisation, licensing transparency and ease of use.

A musician’s test on adamharkus.com runs AI video generators against his own tracks and names Freebeat the winner. The other tools in it are Kaiber, Neural Frames, Runway, Luma Dream Machine and Kling AI. It judges video tools, so it bears on the music-video entries alone.

Renderforest’s comparison of AI music video generators, published by Renderforest, names Renderforest its best overall choice. It says it uses no invented numerical scores. Its rights section covers what a platform’s commercial rights leave out. You stay responsible for any photo, character or third-party footage you upload.

MusicAI’s homepage is a product page. It describes its tracks as royalty-free under a commercial licence. A Reddit thread on music-video tools also ranked, but it blocked the capture and stays unread.

The recorded answers noticed the same pattern. Claude Opus 5 wrote: “Somio’s list picks Somio, MusicGPT’s list picks MusicGPT, Freebeat’s list picks Freebeat.” Perplexity’s answer to the target question cites the adamharkus.com test and the same Reddit thread, so the ranking pages and the AI answers draw on shared sources.

Those guides rank products for purchase. The panel counts which names AI answers produce. It adds two things none of the guides carries: the split by model and a check of the raw answer text against the count.

What these counts cannot tell you

Visibility is in scope and product quality is not. Answer share and named first count presence in recorded answers. A high count means the models name a product often. It says nothing about sound quality, support, licence fairness or fit with your edit. A model can also name a product only to warn against it.

Each model answered each question once, so one unusual answer moves a count by one. The figures describe one edition, 2026-09, and the next edition can differ.

Mentions are matched as text against a list of tracked names. Spellings outside that list go uncounted, which is why Soundraw reads zero for the OpenAI models. Products outside the list get no row at all. Adobe Firefly opens several OpenAI answers and is not tracked in this edition.

The answers were collected programmatically with web search on, and a consumer chat app can answer differently. All five questions are in English. No vendor pays to appear, move or leave the list.

What should a video creator do with this list?

Match the tool to the job, then compare the top two names for that job.

For a song with vocals, start with Suno and compare Udio. For background music under narration, start with Soundraw and compare Beatoven. For long ambient beds or music inside an app, look at Mubert. For orchestral scoring, look at AIVA. For a music video built from a finished track, look at Freebeat or Neural Frames.

Next, read the licence on the plan you would pay for. Suno’s record gives paid subscribers the songs they create. AIVA ties copyright ownership to its Pro plan. Somio’s guide says Soundraw’s paid plans unlock downloads and full commercial usage rights.

Then test the hardest passage. Renderforest advises testing video tools on a section with a transition, such as a beat drop or a vocal entrance. The same test works for a music generator. Ask for the cue your edit needs most and judge the result against it.

If you ask an AI model yourself, expect the answer to depend on the provider. A Perplexity model is likelier to lead with Soundraw. Most other models lead with Suno.

Frequently asked questions

Which AI generator is best for generating music videos?

Freebeat and Neural Frames are the music-video tools the recorded answers name. Freebeat appears in 7 of 50 answers and Neural Frames in 2. The captured guides disagree on the leader. The adamharkus.com test names Freebeat the winner. Renderforest’s comparison names Renderforest best overall, in a guide Renderforest published. Renderforest places Freebeat for performance and lip-sync videos and Neural Frames for audio-reactive work.

Which AI tools are best for video generation?

General video generation sits outside this count, which covers music tools. The adamharkus.com test covers Runway, Luma Dream Machine and Kling AI. It finds that none of them uses your track to drive the video. Renderforest groups Runway and Pika as general generators that create source footage you still cut to the song. For a music-led video, a music-first tool such as Freebeat or Neural Frames removes that editing step.

Is Suno or Udio better?

The counts cannot settle quality, but they show a clear difference in how the models treat the two. Suno is named in 45 of 50 answers and first in 32. Udio is named in 33 and first in none. Somio’s guide separates them by job. It places Suno for DAW-style control and Udio for precise song refinement. For monetised video, Suno’s record ties commercial rights to paid plans, and Gemini’s answer says the same of Udio.

Can I use AI-generated music in monetised videos?

Usually yes, on a paid plan, and the terms differ by product. Suno’s record says songs made by paid subscribers are theirs to keep. AIVA ties full copyright ownership to its Pro plan. Renderforest notes that a platform’s commercial rights do not clear material you upload, such as a celebrity photo or third-party footage. It also notes that YouTube asks creators to disclose realistic AI-generated or altered content.

Why is Adobe Firefly not on the list?

Adobe Firefly was not among the 17 tracked products in this edition, so it has no count. Several OpenAI answers still open with it. ChatGPT’s answer to the target question names Adobe Firefly’s Generate Soundtrack as its overall pick for video creators. Named first counts only tracked products, so in those answers the first tracked name gets the credit.