Media and embeddings

Choose by what goes in and what comes out

Image generation, speech, video, music, segmentation and embeddings have different billing units. Each entry preserves the provider’s units and separates preview access from published prices.

Source checks: October 1, 2026. This is a selected reference of 48 models and endpoints; coverage is incomplete. Prices are USD unless the quote states otherwise. Unknown means a current figure or access contract was not confirmed. Rates, regions, account eligibility and free tiers can change.

Text and reasoning model comparisons

Google

Gemini 3.8 Live

API identifier: gemini-3.8-live

Input: text, image, audio, video
Output: text, audio
Availability: Stable

Provider pricing and conditions

Paid Tier Per Million Tokens / Input Text: $0.75
Paid Tier Per Million Tokens / Input Audio: $3
Paid Tier Per Million Tokens / Input Image Video: $1
Paid Tier Per Million Tokens / Output Text: $4.5
Paid Tier Per Million Tokens / Output Audio: $12
Effective Per Minute / Input Audio: $0.005
Effective Per Minute / Input Image Video: $0.002
Effective Per Minute / Output Audio: $0.018
Grounding: $14/1,000 search requests after the shared 5,000 monthly free allowance

Google quotes both token prices and effective per-minute equivalents for audio/video; do not multiply per-minute values by token-count values or flatten modality rates.

Source check October 1, 2026

Google

Gemini 3.8 Live Extended Thinking

API identifier: gemini-3.8-live-extended-thinking

Input: text, image, audio, video
Output: text, audio
Availability: Stable

Provider pricing and conditions

Same published Live model rates: text input $0.75/MTok; audio input $3/MTok or $0.005/min; image/video input $1/MTok or $0.002/min; text output $4.50/MTok; audio output $12/MTok or $0.018/min

Thinking and asynchronous function calls are supported. Not available for batch; audio/image/video rates are quoted per minute equivalents and token rates in the pricing table. Pricing table explicitly groups Gemini 3.8 Live and Extended Thinking under the same published rates; effective minute equivalents are pricing estimates.

Source check October 1, 2026

Google

Gemini 3.8 Flash TTS

API identifier: gemini-3.8-flash-tts

Input: text
Output: audio
Availability: Endpoint listed; lifecycle label not confirmed

Provider pricing and conditions

Standard Per Million Tokens / Input Text Through 2026-12-31: $0.5
Standard Per Million Tokens / Input Text From 2027-01-01: $1
Standard Per Million Tokens / Output Audio Through 2026-12-31: $9
Standard Per Million Tokens / Output Audio From 2027-01-01: $18
Standard Per Million Tokens / Input Caching Through 2026-12-31: $0.125
Standard Per Million Tokens / Input Caching From 2027-01-01: $0.25
Effective Audio / Per 10 seconds Through 2026-12-31: $0.00225
Effective Audio / Per 10 seconds From 2027-01-01: $0.0045
Batch Per Million / Input Text Through 2026-12-31: $0.25
Batch Per Million / Input Text From 2027-01-01: $0.5
Batch Per Million / Output Audio Through 2026-12-31: $4.5
Batch Per Million / Output Audio From 2027-01-01: $9

Intro rates through Dec 31 2026; rates double Jan 1 2027. Audio output is charged in audio tokens; $/10-second equivalence is a convenience measure. Provider also publishes free-tier entries.

Source check October 1, 2026

Google

Gemini 3.8 Flash-Lite TTS

API identifier: gemini-3.8-flash-lite-tts

Input: text
Output: audio
Availability: Endpoint listed; lifecycle label not confirmed

Provider pricing and conditions

Standard Per Million Tokens / Input Text Through 2026-12-31: $0.5
Standard Per Million Tokens / Input Text From 2027-01-01: $1
Standard Per Million Tokens / Output Audio Through 2026-12-31: $6
Standard Per Million Tokens / Output Audio From 2027-01-01: $12
Standard Per Million Tokens / Input Caching Through 2026-12-31: $0.125
Standard Per Million Tokens / Input Caching From 2027-01-01: $0.25
Effective Audio / Per 10 seconds Through 2026-12-31: $0.0015
Effective Audio / Per 10 seconds From 2027-01-01: $0.003
Batch Per Million / Input Text Through 2026-12-31: $0.25
Batch Per Million / Input Text From 2027-01-01: $0.5
Batch Per Million / Output Audio Through 2026-12-31: $3
Batch Per Million / Output Audio From 2027-01-01: $6

Intro rates through Dec 31 2026; rates double Jan 1 2027. Audio output is charged in audio tokens; $/10-second equivalence is a convenience measure. Provider also publishes free-tier entries.

Source check October 1, 2026

Google

Gemini 3.5 Transcribe

API identifier: gemini-3.5-transcribe

Input: audio (up to 1 hour)
Output: text, word annotations
Availability: Endpoint listed; lifecycle label not confirmed

Provider pricing and conditions

Per Million Tokens / Input Audio: $2
Per Million Tokens / Output Text: $12
Effective Per Minute / Input Audio: $0.003
Effective Per Minute / Output Text: $0.002
Effective Per Minute / Blended Estimate: $0.005

Effective-minute estimate assumes 25 audio tokens/sec and 175 output text tokens/min; cost varies with audio duration and transcript length.

Source check October 1, 2026

Google

Gemini 3.5 Transcribe Live

API identifier: gemini-3.5-transcribe-live

Input: audio
Output: text
Availability: Endpoint listed; lifecycle label not confirmed

Provider pricing and conditions

Per Million Tokens / Input Audio: $3.5
Per Million Tokens / Output Text: $21
Effective Per Minute / Input Audio: $0.005
Effective Per Minute / Output Text: $0.004
Effective Per Minute / Blended Estimate: $0.009

Parent Transcribe card lists this exact identifier as its Live API version. Standalone URL returns 404; the parent card does not publish an independent lifecycle label. Provider effective-minute values assume 25 audio tokens/sec and 175 text tokens/minute.

Source check October 1, 2026

Google

Gemini 3.5 Live Translate

API identifier: gemini-3.5-live-translate-preview

Input: audio speech
Output: audio translated speech, text transcript
Availability: Preview

Provider pricing and conditions

Per Million Tokens / Input Audio: $3.5
Per Million Tokens / Output Audio: $21
Effective Per Minute / Input Audio: $0.0053
Effective Per Minute / Output Audio: $0.0315
Effective Per Minute / Combined Approx: $0.0368

Preview endpoint; provider computes audio-token charges using 25 tokens/sec. Minute value is effective approximate total, not a fixed per-request rate.

Source check October 1, 2026

Google

Nano Banana 2 (Gemini 3.1 Flash Image)

API identifier: gemini-3.1-flash-image

Input: text, image, video, PDF
Output: image, text
Availability: Stable

Provider pricing and conditions

Standard Per Million Tokens / Input Text Image: $0.5
Standard Per Million Tokens / Output Text Thinking: $3
Standard Per Million Tokens / Output Image Tokens: $60
Standard Equivalent Per Image / 0.5K: $0.045
Standard Equivalent Per Image / 1K: $0.067
Standard Equivalent Per Image / 2K: $0.101
Standard Equivalent Per Image / 4K: $0.151
Batch Per Million Tokens / Input Text Image: $0.25
Batch Per Million Tokens / Output Text Thinking: $1.5
Batch Per Million Tokens / Output Image Tokens: $30
Batch Equivalent Per Image / 0.5K: $0.022
Batch Equivalent Per Image / 1K: $0.034
Batch Equivalent Per Image / 2K: $0.05
Batch Equivalent Per Image / 4K: $0.076

Images are billed by output image tokens at $60 per million (standard) with resolution-specific token counts/equivalents. Do not misstate image price as $60 per single image.

Source check October 1, 2026

Google

Nano Banana 2 Lite (Gemini 3.1 Flash-Lite Image)

API identifier: gemini-3.1-flash-lite-image

Input: text, image, video, PDF
Output: image, text
Availability: Stable

Provider pricing and conditions

Standard Per Million Tokens / Input Text Image Video: $0.25
Standard Per Million Tokens / Output Text Thinking: $1.5
Standard Per Million Tokens / Output Image Tokens: $30
Standard Equivalent Per1KImage: $0.0336
Batch Per Million Tokens / Input Text Image Video: $0.125
Batch Per Million Tokens / Output Text Thinking: $0.75
Batch Per Million Tokens / Output Image Tokens: $15
Batch Equivalent Per1KImage: $0.0168

Image output is priced per million image tokens; provider gives 1K resolution equivalent. Keep text-token and generated-image-token units explicit.

Source check October 1, 2026

Google

Nano Banana Pro (Gemini 3 Pro Image)

API identifier: gemini-3-pro-image

Input: image, text
Output: image, text
Availability: Stable

Provider pricing and conditions

Standard Per Million Tokens / Input Text Image: $2
Standard Per Million Tokens / Output Text Thinking: $12
Standard Per Million Tokens / Output Image Tokens: $120
Standard Equivalent Per Image / 1K or 2K: $0.134
Standard Equivalent Per Image / 4K: $0.24
Batch / Input Per Million Text: $1
Batch / Input Per Image: $0.0006
Batch / Output Text Per Million: $6
Batch / Output Image / 1K or 2K: $0.067
Batch / Output Image / 4K: $0.12

The $2 input includes text/image token pricing and Google separately gives a $0.0011/image input equivalent; output image tokens cost $120 per million, with resolution equivalents. Batch rates differ.

Source check October 1, 2026

Google

Nano Banana (Gemini 2.5 Flash Image)

API identifier: gemini-2.5-flash-image

Input: image, text
Output: image, text
Availability: Deprecated; Google says shutdown October 2, 2026

Provider pricing and conditions

Standard / Input Text or Image Per Million Tokens: $0.30
Standard / Output Image Up To 1024x1024: $0.039 per image (equivalent to $30 per 1M image tokens)
Batch / Input Text or Image Per Million Tokens: $0.15
Batch / Output Image: $0.0195 per image
Flex / Input Text or Image Per Million Tokens: $0.15
Flex / Output Image: $0.0195 per image
Priority / Input Text or Image Per Million Tokens: $0.54
Priority / Output Image: $0.0702 per image

Google pricing states this endpoint is deprecated and shuts down 2026-10-02. Output image equivalent depends on 1024x1024 or lower; standard, batch, flex, and priority tiers are distinct.

Source check October 1, 2026

Google

Veo 3.1 Standard

API identifier: veo-3.1-generate-preview

Input: text, image
Output: video with audio
Availability: Endpoint listed; lifecycle label not confirmed

Provider pricing and conditions

Per Second Video With Audio / 720p: $0.4
Per Second Video With Audio / 1080p: $0.4
Per Second Video With Audio / 4K: $0.6

Exact endpoint has literal preview suffix, but the model card does not provide a stable/preview lifecycle label. Publish the literal status text only. The exact endpoint ID includes preview; the inspected parent card does not state a separate lifecycle label.

Source check October 1, 2026

Google

Veo 3.1 Fast

API identifier: veo-3.1-fast-generate-preview

Input: text, image
Output: video with audio
Availability: Endpoint listed; lifecycle label not confirmed

Provider pricing and conditions

Per Second Video With Audio / 720p: $0.1
Per Second Video With Audio / 1080p: $0.12
Per Second Video With Audio / 4K: $0.3

Source check October 1, 2026

Google

Veo 3.1 Lite

API identifier: veo-3.1-lite-generate-preview

Input: text, image
Output: video with audio
Availability: Preview (page title and description); version field does not separately label lifecycle

Provider pricing and conditions

Per Second Video With Audio / 720p: $0.05
Per Second Video With Audio / 1080p: $0.08
Per Second Video With Audio / 4K: Not verified

4K is not supported. Rates are per generated second, not per token or per whole video. The exact endpoint ID includes preview; the inspected parent card does not state a separate lifecycle label.

Source check October 1, 2026

Google

Lyria 3.5

API identifier: lyria-3.5

Input: text, image
Output: audio (MP3), text lyrics
Availability: Stable

Provider pricing and conditions

Per Full Song: $0.08

Source check October 1, 2026

Google

Lyria 3 Clip Preview

API identifier: lyria-3-clip-preview

Input: text, image
Output: audio (MP3), text lyrics
Availability: Preview

Provider pricing and conditions

Per Song: $0.04

Source check October 1, 2026

Google

Lyria Realtime

API identifier: lyria-realtime-exp

Input: weighted text prompts
Output: raw 16-bit PCM audio
Availability: Experimental

Provider pricing and conditions

Not verified

No specific per-request or token price found on the checked Gemini API pricing page; leave unknown rather than borrowing Lyria 3.5 per-song price.

Source check October 1, 2026

Google

Gemini Embedding 2

API identifier: gemini-embedding-2

Input: text, image, video, audio, PDF
Output: text embeddings
Availability: Stable in embedding model-version docs; general model list retains a preview-suffixed alias

Provider pricing and conditions

Unit: USD per 1M modality tokens with per-image/second/frame equivalents
Standard / Text Input Per Million: $0.2
Standard / Image Input Per Million: $0.45
Standard / Image Per Image: $0.00012
Standard / Audio Input Per Million: $6.5
Standard / Audio Per Second: $0.00016
Standard / Video Input Per Million: $12
Standard / Video Per Frame: $0.00079
Batch / Text Input Per Million: $0.1
Batch / Image Input Per Million: $0.225
Batch / Image Per Image: $6e-05
Batch / Audio Input Per Million: $3.25
Batch / Audio Per Second: $8e-05
Batch / Video Input Per Million: $6
Batch / Video Per Frame: $0.000395
Free Tier: Free for standard; batch unavailable on free tier

Google docs differ in model table alias (`gemini-embedding-2-preview`) vs embeddings model-version table and pricing/API stable code (`gemini-embedding-2`). Prefer endpoint model code stated in current version/pricing docs; preserve alias note in evidence. Current pricing uses gemini-embedding-2; the provider model list still uses gemini-embedding-2-preview. Confirm the usable identifier in your account.

Source check October 1, 2026

OpenAI

GPT Realtime 2.1

API identifier: gpt-realtime-2.1

Input: text, image, audio
Output: text, audio
Availability: API available; realtime

Provider pricing and conditions

Unit: USD / 1M modality tokens
Audio Input: $32
Audio Output: $64
Text Input: $4
Text Output: $24
Image Input: $5
Cached Input By Modality / Audio: $0.4
Cached Input By Modality / Text: $0.4
Cached Input By Modality / Image: $0.5

No image output rate is listed; retain null. Model name and pricing are listed in the API price table; a separate lifecycle label is not confirmed here.

Source check October 1, 2026

OpenAI

GPT Realtime 2.1 mini

API identifier: gpt-realtime-2.1-mini

Input: text, image, audio
Output: text, audio
Availability: API available; realtime

Provider pricing and conditions

Unit: USD / 1M modality tokens
Audio Input: $10
Audio Output: $20
Text Input: $0.6
Text Output: $2.4
Image Input: $0.8
Cached Input By Modality / Audio: $0.3
Cached Input By Modality / Text: $0.06
Cached Input By Modality / Image: $0.08

Model name and pricing are listed in the API price table; a separate lifecycle label is not confirmed here.

Source check October 1, 2026

OpenAI

gpt-image-2.5-sunburst

API identifier: gpt-image-2.5-sunburst

Input: text, image
Output: image
Availability: API listed; generation

Provider pricing and conditions

Unit: USD / 1M image tokens; text tokens separately
Image Input: $8
Cached Image Input: $2
Batch Image Input: $4
Batch Cached Image Input: $1
Batch Image Output: $15
Text Cached Input: $1.25
Batch Text Input: $2.5
Batch Cached Text Input: $0.625
Image Output: $30
Text Input: $5
Image Unit Estimate: calculator in image generation guide; not copied here

Standard rates for 2.5; gpt-image-2 standard rates were not present in fetched text (Batch only).

Source check October 1, 2026

OpenAI

gpt-image-2.5-flare

API identifier: gpt-image-2.5-flare

Input: text, image
Output: image
Availability: API listed; generation

Provider pricing and conditions

Unit: USD / 1M image tokens; text tokens separately
Image Input: $8
Cached Image Input: $2
Batch Image Input: $4
Batch Cached Image Input: $1
Batch Image Output: $15
Text Cached Input: $1.25
Batch Text Input: $2.5
Batch Cached Text Input: $0.625
Image Output: $30
Text Input: $5
Image Unit Estimate: calculator in image generation guide; not copied here

Standard rates for 2.5; gpt-image-2 standard rates were not present in fetched text (Batch only).

Source check October 1, 2026

OpenAI

gpt-image-2

API identifier: gpt-image-2

Input: text, image
Output: image
Availability: Current pricing table lists Batch API rates; Standard price and individual model-card lifecycle not verified

Provider pricing and conditions

Unit: USD / 1M image tokens; text tokens separately
Batch Image Input: $4
Batch Cached Image Input: $1
Batch Image Output: $15
Batch Text Input: $2.5
Batch Cached Text Input: $0.625
Standard rates: Not verified
Image Unit Estimate: calculator in image generation guide; not copied here

Standard rates for 2.5; gpt-image-2 standard rates were not present in fetched text (Batch only).

Source check October 1, 2026

OpenAI

gpt-live-transcribe

API identifier: gpt-live-transcribe

Input: audio
Output: text
Availability: API pricing listed

Provider pricing and conditions

Unit: provider token prices and estimated audio-minute cost
Published price / estimated cost: 0.017/minute

Model name and pricing are listed in the API price table; a separate lifecycle label is not confirmed here. The pricing table labels minute costs as estimates; confirm the endpoint’s billing contract before treating an estimate as a fixed minute tariff.

Source check October 1, 2026

Provider source 1

OpenAI

gpt-realtime-whisper

API identifier: gpt-realtime-whisper

Input: audio
Output: text
Availability: API pricing listed

Provider pricing and conditions

Unit: provider token prices and estimated audio-minute cost
Published price / estimated cost: 0.017/minute

Source check October 1, 2026

Provider source 1

OpenAI

gpt-transcribe

API identifier: gpt-transcribe

Input: audio
Output: text
Availability: API pricing listed

Provider pricing and conditions

Unit: USD per minute of audio
Published model-card price: $0.0045/minute

Model name and pricing are listed in the API price table; a separate lifecycle label is not confirmed here. The model card directly states $0.0045 per minute of audio; this is a published model-card price, not a pricing-page estimate.

Source check October 1, 2026

OpenAI

gpt-realtime-translate

API identifier: gpt-realtime-translate

Input: audio
Output: text, audio
Availability: API pricing listed

Provider pricing and conditions

Unit: USD per minute of realtime audio
Published model-card price: $0.034/minute

Model name and pricing are listed in the API price table; a separate lifecycle label is not confirmed here. The model card directly states $0.034 per minute of realtime audio; its output includes translated audio and transcript deltas.

Source check October 1, 2026

OpenAI

gpt-4o-transcribe

API identifier: gpt-4o-transcribe

Input: audio
Output: text
Availability: API pricing listed

Provider pricing and conditions

Unit: provider token prices and estimated audio-minute cost
Published price / estimated cost: 2.50 input + 10 output / MTok; estimate $0.006/min

Source check October 1, 2026

Provider source 1

OpenAI

gpt-4o-mini-transcribe

API identifier: gpt-4o-mini-transcribe

Input: audio
Output: text
Availability: API pricing listed

Provider pricing and conditions

Unit: provider token prices and estimated audio-minute cost
Published price / estimated cost: 1.25 input + 5 output / MTok; estimate $0.003/min

Source check October 1, 2026

Provider source 1

OpenAI

text-embedding-3-small

API identifier: text-embedding-3-small

Input: text
Output: vector embedding
Availability: OpenAI API model card; active catalog entry

Provider pricing and conditions

Unit: USD / 1M tokens
Input: $0.02

Text embedding only; do not describe as multimodal embedding.

Source check October 1, 2026

OpenAI

text-embedding-3-large

API identifier: text-embedding-3-large

Input: text
Output: vector embedding
Availability: OpenAI API model card; active catalog entry

Provider pricing and conditions

Unit: USD / 1M tokens
Input: $0.13

Text embedding only; do not describe as multimodal embedding.

Source check October 1, 2026

Muse Image

API identifier: muse-image-1.0

Input: text, image
Output: image
Availability: Hosted Meta Model API; available

Provider pricing and conditions

Unit: USD / generated image
Price: $0.01

Image-search included; billed only for successful returned images.

Source check October 1, 2026

Muse Voice Transcribe

API identifier: muse-voice-transcribe-1.0

Input: audio
Output: text
Availability: Hosted Meta Model API; available

Provider pricing and conditions

Unit: USD / audio hour
Price: $0.18

Streaming and file transcription same rate; no contributor tier.

Source check October 1, 2026

Segment Anything Model 3.1

API identifier: sam-3.1

Input: text, image, video
Output: segmentation masks, boxes
Availability: Hosted Meta Model API; available

Provider pricing and conditions

Unit: USD / usage unit
Image Per 1,000: $2.5
Video Frame Per 1,000: $0.2

Image and video-frame units; not token rates.

Source check October 1, 2026

Amazon

Nova Multimodal Embeddings

API identifier: amazon.nova-2-multimodal-embeddings-v1:0

Input: text, image, video
Output: vector embeddings
Availability: Amazon Bedrock model lifecycle Active; launch date 2025-10-28; model card says EOL no sooner than 2026-10-28

Provider pricing and conditions

Not verified

AWS describes this embedding model as converting text, images, and video to vectors; its modality table is ambiguous about additional inputs, so only the explicitly described modalities are listed. Region-specific rates remain unverified.

Source check October 1, 2026

Amazon

Nova 2 Sonic

API identifier: amazon.nova-2-sonic-v1:0

Input: speech audio (exact input format details not specified in model card)
Output: speech/audio (exact output format details not specified in model card)
Availability: Amazon Bedrock model lifecycle Active; launch date 2025-12-02; model card says EOL no sooner than 2026-12-02

Provider pricing and conditions

Not verified

AWS describes Nova 2 Sonic as speech-to-speech; the model card’s modality table appears to mix modalities and API families, so only the explicit speech-to-speech capability is reported. Price remains unknown. AWS’s model card lists 1M-token context and a 64K output cap; these are model limits, distinct from audio duration.

Source check October 1, 2026

Microsoft

MAI Image 2.6

API identifier: Not verified; confirm the deployment identifier

Input: text, image (up to 5 images for editing)
Output: one image
Availability: Microsoft Foundry Preview; version 2026-07-31; Global Standard deployment; image generation and editing

Provider pricing and conditions

Not verified

Microsoft Foundry documents 32,000-token request context, up to five JPEG/PNG editing references and one PNG output. Each image dimension must be at least 768; width × height must not exceed 2,359,296 pixels. Either dimension can exceed 1536 if the total stays under the cap. Supported request options include width, height, prompt, auto_aspect_ratio and web_grounding. The provider product page’s Arena rank claim is undated and is not used as a verified ranking here. Current price and deployment identifier remain unverified.

Source check October 1, 2026

Microsoft

MAI Image 2.6 Flash

API identifier: Not verified; confirm the deployment identifier

Input: text, image (up to 5 images for editing)
Output: one image
Availability: Microsoft Foundry Preview; version 2026-07-31; Global Standard deployment; image generation and editing

Provider pricing and conditions

Not verified

Current per-unit price and deployment identifier must be checked by cloud/region; display name is not confirmed API ID.

Source check October 1, 2026

Microsoft

MAI Transcribe 2

API identifier: Not verified; confirm the deployment identifier

Input: audio
Output: text
Availability: Microsoft Azure Speech public preview; public preview carries no SLA and is not recommended for production; current supported list includes MAI-Transcribe-2/1.5; MAI-Transcribe-1 deprecated 2026-08-20

Provider pricing and conditions

Limited time: USD $0.10 per hour of audio (Microsoft AI model page; not a token rate).

Microsoft lists 60 languages, diarization and word-level timestamps. Its product page reports 3.4% average WER on FLEURS’s top 25 languages and 2% WER on Artificial Analysis. These are provider-page reports; exact benchmark run dates, configurations and publisher snapshots are not established here. The $0.10/audio-hour offer is limited-time with no verified end date; confirm region/account billing. Deployment identifier remains unverified.

Source check October 1, 2026

OpenAI

GPT-4o Mini TTS

API identifier: gpt-4o-mini-tts

Input: text
Output: audio
Availability: Current OpenAI API model card; undated alias and dated snapshots listed

Provider pricing and conditions

Unit: USD / 1M tokens
Text Input: $0.6
Audio Output: $12

Dated snapshots include 2025-12-15 and 2025-03-20; current token pricing, not per-minute or per-character.

Source check October 1, 2026

OpenAI

GPT Audio 1.5

API identifier: gpt-audio-1.5

Input: text, audio
Output: text, audio
Availability: Current OpenAI API model card; undated alias and dated snapshots listed

Provider pricing and conditions

Unit: USD / 1M tokens
Text Input: $2.5
Text Output: $10
Audio Input: $32
Audio Output: $64

Model card describes generally available audio model; text/audio input and output; no image/video.

Source check October 1, 2026

ByteDance Seed

Seedance 2.5

API identifier: Not verified; confirm the deployment identifier

Input: text prompts plus up to 30 images, 10 video clips, and 10 audio clips as references per pass
Output: audio-video clip
Availability: Official launch post dated 2026-07-31 says rolling out on Jimeng AI and Doubao Pro; BytePlus ModelArk API access was coming soon as of that post. Current official-source search did not confirm a later API launch.

Provider pricing and conditions

Not verified in captured provider sources

The July 31 launch post reports generation up to 30 seconds in one pass and multi-round extension, with no fixed total-duration cap documented in this review. The post announced platform rollout and future ModelArk API access; current API availability, exact model ID, prices, weights/license and benchmark setup remain unverified.

Source check October 1, 2026

ByteDance Seed

Seedream 5.0 Pro

API identifier: Not verified; confirm the deployment identifier

Input: text prompts and image references for editing are evidenced by the official product page examples; exact API input contract unknown
Output: generated image
Availability: Listed as a current ByteDance Seed model; official page does not establish an API endpoint or precise access route

Provider pricing and conditions

Not verified in captured provider sources

Official product examples demonstrate image generation and editing. Exact serving input/output contract, release date, image dimensions, access route, prices, weights/license and benchmark setup remain unverified.

Source check October 1, 2026

xAI

Grok Imagine Image 2.0

API identifier: grok-imagine-image-2.0

Input: text, image
Output: image
Availability: API listed; account and region availability apply

Provider pricing and conditions

Image input $0.01/image
Low output: 1K $0.04/image, 1.5K $0.05/image, 2K $0.06/image
Medium output: 1K $0.06/image, 1.5K $0.07/image, 2K $0.08/image

Quality auto currently chooses low for generation and medium for editing; billing follows the quality served. Batch supported without a listed discount. November 2 migration from the older quality alias defaults to low; see exact migration notice.

Source check October 1, 2026

xAI

Grok Imagine Image

API identifier: grok-imagine-image

Input: text, image
Output: image
Availability: API listed; account and region availability apply

Provider pricing and conditions

Image input $0.002/image
Output $0.02/image at 1K or 2K

Batch supported in the listed US regions without a listed discount. The November 2 quality-alias migration does not affect this model.

Source check October 1, 2026

xAI

Grok Imagine Image Quality

API identifier: grok-imagine-image-quality

Input: text, image
Output: image
Availability: API listed; account and region availability apply

Provider pricing and conditions

Image input $0.01/image
Output: 1K $0.05/image, 1.5K $0.06/image, 2K $0.07/image

Scheduled API retirement November 2, 2026: the slug continues resolving but requests redirect to image-2.0 at low quality and its rates. This is a future serving change, not an October 1 shutdown. Batch not supported.

Source check October 1, 2026

xAI

Grok Imagine Video 1.5

API identifier: grok-imagine-video-1.5

Input: text, image, audio (preset voices)
Output: video
Availability: API listed; account and region availability apply

Provider pricing and conditions

Image input $0.01/image
Preset-voice audio input free
Output: 480p $0.08/second, 720p $0.14/second, 1080p $0.25/second

Output billing is per generated video second. Batch supported in listed US regions without a listed discount; account/region limits apply.

Source check October 1, 2026

xAI

Grok Imagine Video

API identifier: grok-imagine-video

Input: text, image, video
Output: video
Availability: API listed; account and region availability apply

Provider pricing and conditions

Image input $0.002/image
Video input $0.01/second
Output: 480p $0.05/second, 720p $0.07/second

Video extension prices are promotional and subject to change. Batch supported in listed US regions without a listed discount; do not infer the 1.5 audio-input contract for this model.

Source check October 1, 2026