I really hope it'll be made free for STM32 too.
ST is actively making their toolsuite sexy and up to date, looking forward to the future.
Same! STM32 is quickly becoming my favorite MCU. Not only the software part is better overall (there's a bunch of other cool things like audio weaver, or X-CUBE-VS4A if you're into that alexa stuff), but the hardware itself is great. I've used several other ARM Cortex M MCUs (LPC Series, Kinetis, MSP432, Tiva C...) and STM32 is quickly getting ahead of the pack. My main MCU for the last while has been the Tiva C series which is an M4F but it's getting a bit limited for what I do. The STM32 has M7 parts, and price-wise its a much better value too.
Here's a simple, quick comparison between my "usual" TM4C1294NCPDT and the STM32F769BI, which does cost 10% more:
-M4F @ 120MHz vs M7 @ 216MHz; the M7 is about 50% faster at the same clock speed to begin with so it's a big gain (also supports double precision FP)
-1MB vs 2MB flash
-256KB vs 512KB RAM
-90 vs 140 GPIO -- more GPIOs which are also faster and more which are 5V tolerant
-20x12b (2x 2MSPS) vs 24x12b ADC (3x 2.4MSPS)
-2x CAN 2.0B bus vs 3x CAN 2.0B bus + also LIN (for automotive-style networking; STM32H7 also gets CAN-FD)
-1x USB OTG vs 2x USB OTG
-8x UART vs 8x U(S)ART
-10x I2C vs 4x -- the only place STM32 loses, but being a shared bus it's not a big deal IMO
-4x QSPI vs 6x QSPI which for flash also supports DDR transfers + dual flash so much faster, and also XIP
-both have 10/100 ethernet
-both have similar 16/32 bit timer features
-nothing vs 2x 12b DAC
-nothing vs MMC/SD/SDIO interfaces
-nothing vs 3x I2S and 2x SAI (supports PCM, TDM, DFSDM, SPDIF, AC97...) for audio
-nothing vs LCD interface + MIPI DSI interface (for "mobile" LCDs, can also support HDMI) + Chrom-ART accelerator (with DMA) + HW JPEG Codec for HMIs (with free STEmWin and maybe TouchGFX soon too)
-nothing vs support for various "consumer" protocols like HDMI-CEC, DCMI for cameras, etc.
So: more than double the speed, double the flash, double the RAM, ~50% more GPIOs, better peripherals overall, and it absolutely destroys the Tiva C when it comes to audio or graphics... The benefits are very clear too. The faster core better for real-time DSP work (all those biquads!), more memory is useful for ping-pong buffers (for input/output via DMA)/processing FFTs/frame buffers, etc. I also prefer atollic truestudio to CCS, and CubeMX to PinMux... HAL isn't perfect, but neither is Tivaware (things like different clock functions for different devices e.g. SysCtlClockGet working on TM4C123x but not on TM4C129x) Oh, and the USB DFU mode is nice too (vs having to use LMFlash for updates via USB )