The latest AI news and updates of Pika — AI video company building an idea-to-video platform with generation and editing tools. Covering Pika's latest product updates and launches from the past 90 days.
PikaSep 9Pika added OpenAI's GPT Image 2.5 to the Pika API Club, featuring two new models: Sunburst for precision-quality editing and Flare for faster generations. Both models support optional transparent backgrounds and are available for $30 per million tokens. The integration provides image-to-image and text-to-image capabilities through a single API key and consistent request shape.
PikaSep 3Pika added Google's Gemini Omni 1.1 Flash model to the Pika API Club. The membership program provides access to over 100 generative media models through a single API at wholesale-level pricing. Members pay a $10 monthly fee to access these models on a pay-as-you-go basis, bypassing the markups found on other API aggregation platforms.
PikaAug 26Pika launched WAN 3.0 Prime, a variant of the WAN 3.0 video model that maintains original output quality while delivering 7x faster generation times. The model supports 20 references, enhanced visual and audio realism, and 30-second single-take generations. Access is available through the Pika API Club, which offers pricing up to 35% lower than competing video generation APIs.
PikaAug 25Pika added Alibaba's WAN 3.0 video generation model to the Pika API Club. The model supports up to 20 references, enhanced visual and audio realism, and native 30-second single-take generations. Access is available through the Pika API Club at pricing up to 35% less expensive than competing video generation APIs.
PikaAug 22Pika launched Pika Speech, a 3B-parameter text-to-speech model that generates one minute of 48 kHz studio-quality audio in about 1.2 seconds. The model supports requests up to five minutes and costs $0.01 per minute, which Pika reports is up to 9× more cost-efficient than ElevenLabs v3. Pika Speech is available now through the Pika API Club.
PikaAug 22Pika launched Pika Music, a model that generates complete tracks from text, lyrics, voice references, and music references. The system resolves these inputs within a single shared latent-diffusion decoder, avoiding chained workflows. It generates 90-second songs in 6.21 seconds on average and offers up to 10x greater cost efficiency than other music models, available via the Pika API Club.
PikaAug 22Pika reports its SFX text-to-audio model achieves an average end-to-end generation time of 0.847 seconds, significantly faster than the 2.47-second average of competing hosted APIs. The model is priced up to 95% lower than alternatives and is available now through the Pika API Club.
PikaAug 22Pika released Soundtrack, a video-to-audio model that generates synchronized music, voiceover, and motion-aware sound effects for existing footage. The model ranked first for semantic alignment and audiovisual synchronization on a 67-chunk benchmark. Soundtrack is available through the Pika API Club at pricing up to 50% lower than competing models.
PikaAug 22Pika Labs added the Seedance 2.5 video generation model in 1080p to the Pika API Club. The model delivers sharper textures and cleaner details, with pricing up to 60% cheaper than competitors. API Club members access the model through a unified API for pay-as-you-go video generation.
PikaAug 15Pika launched four generative audio foundation models—Soundtrack, Music, SFX, and Speech—on its API Club. These models offer up to 20x better cost-efficiency than existing market alternatives, with pricing starting at $0.0002 per second. Access is exclusive to API Club members, providing a unified API for video-to-audio, text-to-audio, and music generation tasks.
PikaAug 8Pika added ByteDance's Seedance 2.5 video generation model to the Pika API Club. Members access the model at permanent wholesale pricing, which is up to 88% cheaper than rates found on other API aggregation platforms. The API Club requires a $10 monthly membership fee and provides pay-as-you-go access to over 100 generative media models.
PikaAug 5Pika added the Flux 3 generative model to the Pika API Club. The membership program provides access to over 100 generative media models at wholesale-level pricing through a single API. Members pay a $10 monthly fee to access these models at pay-as-you-go rates, bypassing the markups found on other API aggregation platforms.
PikaAug 5Pika launched the Pika API Club, a $10/month membership providing access to over 100 generative media models through a single API. The program offers wholesale-level pricing, with discounts of up to 87% on models like Seedance 2.0, MiniMax H3, and GPT Image 2 compared to other aggregators. New members receive a $10 credit toward their first month of usage.
PikaJul 31Pika added the MiniMax H3 multimodal video model to the Pika MCP. This integration enables AI agents to generate video by combining up to nine images, three video clips, and three audio clips as references. The model produces native stereo audio and maintains subject, motion, and audio consistency across generations.
PikaJul 11Pika integrated Google’s Gemini Omni model into its Model Context Protocol connector. This update enables AI agents to perform complex video edits, including changing backgrounds, camera angles, and outfits, or adding visual effects and dubbing speech into other languages, through simple natural language prompts.
PikaJul 10Pika launched the 4K-VFX Skill, which transforms video clips using simple prompts while preserving original faces, gestures, audio, and camera movements. Powered by the Seedance 4k model, the skill is available now on the PikaMCP and can be triggered using the /4k-vfx command.
PikaJul 2Pika launched the Voxel-It skill, which converts uploaded photos into Minecraft-style voxel worlds. The tool voxelizes backgrounds and objects while keeping human subjects fully realistic. The skill is also accessible via the Pika MCP for integration into AI agent workflows.
PikaJun 25Pika now offers native 4K resolution for the Seedance 2.0 video generation model through its Model Context Protocol (MCP) connector. This update enables AI agents connected to the Pika MCP to generate high-fidelity video content directly within existing interfaces. Installing the latest companion skills via the Pika MCP server enables 4K generation.
PikaJun 15Pika launched Director's Suite, an experimental interface for end-to-end video creation. Powered by Claude, the platform uses an AI agent to manage concepts, storyboards, and clip generation through chat or voice commands. The agent maintains persistent project context across the timeline, allowing for iterative editing without re-explaining project details. Pika demonstrated the tool with a 6-minute TV pilot.
PikaJun 15Pika launched a Language Swap skill for its Model Context Protocol connector. The tool translates the spoken language in any video while preserving the original speaker's voice and syncing lip movements to the new language. The skill runs through Pika's MCP server, processing video content directly within existing AI agent interfaces.
Frequently asked questions